Ecology of the Shortgrass Steppe: A Long-Term Perspective [First Edition] 0195135822, 9780195135824, 9780199722808

Table of contents :
Contents......Page 8
Contributors......Page 12
1 The Shortgrass Steppe: The Region and Research Sites......Page 18
2 Climate of the Shortgrass Steppe......Page 29
3 Soil Development and Distribution in the Shortgrass Steppe Ecosystem......Page 45
4 Land-Use History on the Shortgrass Steppe......Page 70
5 Vegetation of the Shortgrass Steppe......Page 85
6 The Role of Disturbances in Shortgrass Steppe Community and Ecosystem Dynamics......Page 99
7 Simulation of Disturbances and Recovery in Shortgrass Steppe Plant Communities......Page 134
8 Ecology of Mammals of the Shortgrass Steppe......Page 147
9 Birds of the Shortgrass Steppe......Page 196
10 Insect Populations, Community Interactions, and Ecosystem Processes in the Shortgrass Steppe......Page 230
11 Trophic Structure and Nutrient Dynamics of the Belowground Food Web within the Rhizosphere of the Shortgrass Steppe......Page 263
12 Net Primary Production in the Shortgrass Steppe......Page 285
13 Soil Organic Matter and Nutrient Dynamics of Shortgrass Steppe Ecosystems......Page 321
14 Soil–Atmosphere Exchange of Trace Gases in the Colorado Shortgrass Steppe......Page 357
15 The Shortgrass Steppe and Ecosystem Modeling......Page 388
16 Effects of Grazing on Vegetation......Page 404
17 Cattle Grazing on the Shortgrass Steppe......Page 462
18 Effects of Grazing on Abundance and Distribution of Shortgrass Steppe Consumers......Page 474
19 The Future of the Shortgrass Steppe......Page 499
B......Page 526
C......Page 528
D......Page 529
G......Page 530
H......Page 531
M......Page 532
P......Page 533
R......Page 534
S......Page 535
T......Page 536
Y......Page 537

Citation preview

Ecology of the Shortgrass Steppe

LONG-TERM ECOLOGICAL RESEARCH NETWORK SERIES LTER Publications Committee Grassland Dynamics: Long-Term Ecological Research in Tallgrass Prairie Editors: Alan K. Knapp, John M. Briggs, David C. Hartnett, and Scott L. Collins Standard Soil Methods for Long-Term Ecological Research Editors: G. Philip Robertson, David C. Coleman, Caroline S. Bledsoe, and Phillip Sollins Structure and Function of an Alpine Ecosystem: Niwot Ridge, Colorado Editors: William D. Bowman and Timothy R. Seastedt Climate Variability and Ecosystem Response at Long-Term Ecological Sites Editors: David Greenland, Douglas G. Goodin, and Raymond C. Smith Biodiversity in Drylands: Toward a Unified Framework Editors: Moshe Shachak, James R. Gosz, Steward T. A. Pickett, and Avi Perevolotsky Long-Term Dynamics of Lakes in the Landscape: Long-Term Ecological Research on North Temperate Lakes Editors: John J. Magnuson, Timothy K. Kratz, and Barbara J. Benson Alaska’s Changing Boreal Forest Editors: F. Stuart Chapin, III, Mark W. Oswood, Keith Van Cleve, Leslie A. Viereck, and David L. Verbyla Structure and Function of a Chihuahuan Desert Ecosystem: The Jornada Basin Long-Term Ecological Research Site Editors: Kris M. Havstad, Laura F. Huenneke, and William H. Schlesinger Principles and Standards for Measuring Primary Production Editors: Timothy J. Fahey and Alan K. Knapp Agrarian Landscapes in Transition: Comparisons of Long-Term Ecological and Cultural Change Editors: Charles L. Redman and David R. Foster Ecology of the Shortgrass Steppe: A Long-Term Perspective Editors: William K. Lauenroth and Ingrid C. Burke

Ecology of the Shortgrass Steppe A Long-Term Perspective Edited by WILLIAM K. LAUENROTH INGRID C. BURKE

1 2008

3 Oxford University Press, Inc., publishes works that further Oxford University’s objective of excellence in research, scholarship, and education. Oxford New York Auckland Cape Town Dar es Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico City Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam

Copyright © 2008 by Oxford University Press, Inc. Published by Oxford University Press, Inc. 198 Madison Avenue, New York, New York 10016 www.oup.com Oxford is a registered trademark of Oxford University Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press. Library of Congress Cataloging-in-Publication Data Ecology of the shortgrass steppe : a long-term perspective / edited by William K. Lauenroth and Ingrid C. Burke. p. cm. Includes bibliographical references and index. ISBN 978-0-19-513582-4 1. Grassland ecology—North America. 2. Steppe ecology—North America. I. Lauenroth, William K. II. Burke, Ingrid C. QH102.S56 2009 577.4’4097—dc22 2008006014

9 8 7 6 5 4 3 2 1 Printed in the United States of America on acid-free paper

Acknowledgments

This book is the result of more than four decades of research and represents the efforts of many dedicated individuals, some authors and some not. In many significant ways, Dr. George Van Dyne began the program that continues today as the Shortgrass Steppe Long-Term Ecological project; his vision permeates our presentation. Countless field, laboratory, and program assistants and graduate students have contributed their hard work and insight to our understanding of the shortgrass steppe, and this book is richer as a result. Special thanks to Bob Flynn, Judy Hendryx, Nicole Kaplan, Mark Lindquist, Kim Melville-Smith, Jeri Morgan, Sallie Sprague, Petra Lowe, Sonia Hall, Mark Gathany, and especially Becky Riggle, who assisted in the final preparation of the manuscript. To all of you and more—many thanks.

v

This page intentionally left blank

Contents

Contributors 1

xi

The Shortgrass Steppe The Region and Research Sites 3 William K. Lauenroth, Ingrid C. Burke, and Jack A. Morgan

2

Climate of the Shortgrass Steppe

14

Roger A. Pielke, Sr., and Nolan J. Doesken 3

Soil Development and Distribution in the Shortgrass Steppe Ecosystem 30 Eugene F. Kelly, Caroline M. Yonker, Steve W. Blecker, and Carolyn G. Olson

4

Land-Use History on the Shortgrass Steppe 55 Richard H. Hart

5

Vegetation of the Shortgrass Steppe

70

William K. Lauenroth 6

The Role of Disturbances in Shortgrass Steppe Community and Ecosystem Dynamics 84 Debra P. C. Peters, William K. Lauenroth, and Ingrid C. Burke

7

Simulation of Disturbances and Recovery in Shortgrass Steppe Plant Communities 119 Debra P. C. Peters, and William K. Lauenroth vii

viii

Contents

8 Ecology of Mammals of the Shortgrass Steppe

132

Paul Stapp, Beatrice Van Horne, and Mark D. Lindquist 9 Birds of the Shortgrass Steppe

181

John A. Wiens, and Nancy E. McIntyre 10 Insect Populations, Community Interactions,

and Ecosystem Processes in the Shortgrass Steppe 215 Thomas O. Crist 11 Trophic Structure and Nutrient Dynamics of the Belowground Food Web within the Rhizosphere of the Shortgrass Steppe 248

John C. Moore, Jill Sipes, Amanda A. Whittemore-Olson, H. William Hunt, Diana H. Wall, Peter C. de Ruiter, and David C. Coleman 12 Net Primary Production in the Shortgrass Steppe

270

William K. Lauenroth, Daniel G. Milchunas, Osvaldo E. Sala, Ingrid C. Burke, and Jack A. Morgan 13 Soil Organic Matter and Nutrient Dynamics of Shortgrass Steppe Ecosystems 306

Ingrid C. Burke, Arvin R. Mosier, Paul B. Hook, Daniel G. Milchunas, John E. Barrett, Mary Ann Vinton, Rebecca L. McCulley, Jason P. Kaye, Richard A. Gill, Howard E. Epstein, Robin H. Kelly, William J. Parton, Caroline M. Yonker, Petra Lowe, and William K. Lauenroth 14 Soil–Atmosphere Exchange of Trace Gases in the Colorado Shortgrass Steppe 342

Arvin R. Mosier, William J. Parton, Roberta E. Martin, David W. Valentine, Dennis S. Ojima, David S. Schimel, Ingrid C. Burke, E. Carol Adair, and Stephen. J. Del Grosso 15 The Shortgrass Steppe and Ecosystem Modeling

William J. Parton, Stephen J. Del Grosso, Ingrid C. Burke, and Dennis S. Ojima 16 Effects of Grazing on Vegetation

389

Daniel G. Milchunas, William K. Lauenroth, Ingrid C. Burke, and James K. Detling

373

Contents ix

17 Cattle Grazing on the Shortgrass Steppe

447

Richard H. Hart and Justin D. Derner 18 Effects of Grazing on Abundance and Distribution of Shortgrass Steppe Consumers 459

Daniel G. Milchunas and William K. Lauenroth 19 The Future of the Shortgrass Steppe

484

Ingrid C. Burke, William K. Lauenroth, Michael F. Antolin, Justin D. Derner, Daniel G. Milchunas, Jack A. Morgan, and Paul Stapp Index

511

This page intentionally left blank

Contributors

E. Carol Adair, Department of Forest Sciences, Graduate Degree Program in Ecology Colorado State University, Fort Collins, CO 80523

Peter C. de Ruiter, Alterra, Wageningen University and Research Centre, Wageningen, The Netherlands

Michael F. Antolin, Department of Biology, Colorado State University, Fort Collins, CO 80523

Justin D. Derner, U.S. Department of Agriculture–Agricultural Research Service, High Plains Grasslands Research Station, Cheyenne, WY 82009

John E. Barrett, Department of Biological Sciences, Virginia Tech, Blacksburg, VA 24061

James K. Detling, Department of Biology, Colorado State University, Fort Collins, CO 80523

Steve W. Blecker, U.S. Geological Survey, Mackay School of Earth Science and Engineering, University of Nevada, Reno, NV 89557

Nolan J. Doesken, Atmospheric Science Department and Colorado Climate Center, Colorado State University, Fort Collins, CO 80523

Ingrid C. Burke, Graduate Degree Program in Ecology, Colorado State University, Fort Collins, CO 80523

Howard E. Epstein, Environmental Science Department, University of Virginia, Charlottesville, VA 22904

David C. Coleman, Institute of Ecology, University of Georgia, Athens, GA 30602 Thomas O. Crist, Department of Zoology, Miami University, Oxford, OH 45056

Richard A. Gill, School of Earth and Environmental Sciences, Washington State University, Pullman, WA 99164

Stephen J. Del Grosso, U.S. Department of Agriculture– Agricultural Research Service, Soil Plant Nutrient Research, Fort Collins, CO 80526

Richard H. Hart, U.S. Department of Agriculture–Agricultural Research Service, High Plains Grasslands Research Station, Cheyenne, WY 82009

xi

xii

Contributors

Paul B. Hook, Intermountain Aquatics, Inc., Driggs, ID 83422 H. William Hunt, Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO 80524 Jason P. Kaye, Department of Crop and Soil Sciences, Penn State University, University Park, PA 16802 Eugene F. Kelly, Department of Soil and Crop Sciences, Colorado State University, Fort Collins, CO 80523 Robin H. Kelly, Colorado State University, Fort Collins, CO 80523 William K. Lauenroth, Graduate Degree Program in Ecology, Warner College of Natural Resources, Colorado State University, Fort Collins, CO 80523 Mark D. Lindquist, Shortgrass Steppe Long-Term Ecological Research, Colorado State University, Fort Collins, CO 80523 Petra Lowe, Shortgrass Steppe Long-Term Ecological Research, Colorado State University, Fort Collins, CO 80523 Roberta E. Martin, Department of Global Ecology, Carnegie Institution, Stanford, CA 94305 Rebecca L. McCulley, Plant and Soil Sciences, University of Kentucky, Lexington, KY 40546 Nancy E. McIntyre, Department of Biological Sciences, Texas Tech University, Lubbock, TX 79409

Daniel G. Milchunas, Department of Forest, Rangeland, and Watershed Stewardship, Colorado State University, Fort Collins, CO 80523 John C. Moore, Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO 80524 Jack A. Morgan, U.S. Department of Agriculture–Agricultural Research Service, Fort Collins, CO 80526 Arvin R. Mosier, U.S. Department of Agriculture–Agricultural Research Service, Soil–Plant–Nutrient Research Unit, Fort Collins, CO 80521 Dennis S. Ojima, Colorado State University, Natural Resource Ecology Laboratory, Fort Collins, CO 80523 Carolyn G. Olson, U.S. Department of Agriculture–Natural Resource Conservation Service, Lincoln, NE 68508 William J. Parton, Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO 80523 Debra P. C. Peters, U.S. Department of Agriculture–Agricultural Research Service, Jornada Exp. Range, Las Cruces, NM 88003 Roger A. Pielke, Sr., Cooperative Institute for Research in Environmental Sciences and Department of Atmospheric and Ocean Sciences, Boulder, CO 80309

Contributors xiii

Osvaldo E. Sala, Center for Environmental Studies, Brown University, Providence, RI 02912

Mary Ann Vinton, Department of Biology, Creighton University, Omaha, NE 68178

David S. Schimel, Terrestrial Sciences/Climate and Global Dynamics Division, National Center for Atmospheric Research, Boulder, CO 80307

Diana H. Wall, Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, CO 80524

Jill Sipes, School of Biological Sciences, University of Northern Colorado, Greeley, CO 80639 Paul Stapp, Department of Biological Science, California State University Fullerton, Fullerton, CA 92834 David W. Valentine, Department of Forest Sciences, University of Alaska, Fairbanks, AK 99775 Beatrice Van Horne, U.S. Department of Agriculture Forest Service, Wildlife, Fish, Watershed, and Air Research, Arlington, VA 22209

Amanda A. Whittemore-Olson, School of Biological Sciences, University of Northern Colorado, Greeley, CO 80639 John A. Wiens, The Nature Conservancy, Arlington, VA 22203 Caroline M. Yonker, Department of Soil and Crop Science, Colorado State University, Fort Collins, CO 80523

This page intentionally left blank

Ecology of the Shortgrass Steppe

This page intentionally left blank

1 The Shortgrass Steppe The Region and Research Sites William K. Lauenroth Ingrid C. Burke Jack A. Morgan

On the shortgrass prairie, the green hills roll away toward the distant horizon, uncluttered by buildings, signs, or paved roads. Pronghorn herds graze peacefully while prairie dogs chatter from their burrows. At night coyotes howl, horned owls call, and poorwills sing, just as they have for thousands of years. The land seems eternal. —Cushman and Jones (1988, p. 9)

The central grassland region of North America (Fig. 1.1) is the largest contiguous grassland environment on earth. Prior to European settlement, it was a vast, treeless area characterized by dense head-high grasses in the wet eastern portion, and very short sparse grasses in the dry west. As settlers swept across the area, they replaced native grasslands with croplands, most intensively in the east, and less so in the west (Fig. 1.2). The most drought-prone and least productive areas have survived as native grasslands, and the shortgrass steppe occupies the warmest, driest, least productive locations. James Michener (1974) provided an apt description of the harshness of the shortgrass region in his book Centennial: It is not a hospitable land, like that farther east in Kansas or back near the Appalachians. It is mean and gravelly and hard to work. It lacks an adequate topsoil for plowing. It is devoid of trees or easy shelter. A family could wander for weeks and never find enough wood to build a house. (p. 64)

The objective of this chapter is to introduce the shortgrass steppe (Fig. 1.3) and its record of ecological research. First we present an ecological history of the shortgrass steppe since the Tertiary, and provide the geographic and climatic context for the region. Second we describe the major research sites, and the history of 3

4

Ecology of the Shortgrass Steppe

N

Northern Mixed Shortgrass Steppe Southern Mixed Tallgrass Prairie

250

0

250

500

Kilometers

Figure 1.1 Map of the central grassland region of North America (Lauenroth et al., 1999).

the three major entities or programs that have shaped much of the science done in the shortgrass steppe: the U.S. Department of Agriculture (USDA)–Agricultural Research Service (ARS), the International Biological Programme (IBP), and the Long-Term Ecological Research (LTER) Program.

The Shortgrass Steppe Grasses have been an important component of the shortgrass steppe of North America since the Miocene (5–24 million years ago) (Axelrod, 1985; Stebbins, 1981). Before that, during the Paleocene and Eocene (34–65 million years ago), the vegetation was a mixture of temperate and tropical mesophytic forests. Two causes have been proposed as explanations for this ancient change from forest

The Shortgrass Steppe: The Region and Research Sites 5

Figure 1.2

Current land use in the shortgrass steppe of North America (NLCD, 2001).

to grassland. First, global temperatures decreased rapidly during the Oligocene (24–34 million years ago), creating conditions for a drier climate. These drier conditions, combined with a renewal of the uplift of the Rocky Mountains that had begun during the Paleocene, left the Great Plains in a rain shadow. Thus, the Plains became dry and cool, causing a shift from forest to grassland vegetation. Currently, the vegetation is dominated by C4 grasses, a characteristic of many

6

Ecology of the Shortgrass Steppe

Figure 1.3 Shortgrass steppe landscape with the Ogallala escarpment in the background. (Photo by Sallie Sprague.)

world grasslands today that is attributed to the relatively low atmospheric CO2 concentrations that have prevailed from the late Miocene up until the early 20th century and that tend to favor C4 over C3 metabolism (Ehleringer et al., 1997). After the rise to dominance by grasses in the Great Plains, there were four major episodes of continental glaciation during the Pleistocene, beginning with the Nebraskan glacial period and terminating with the Wisconsin, which ended 10,000 years ago. Separating these glacial periods were warmer interglacial periods. This glacial/interglacial alternation lasted for a million years and had a substantial impact on the region. The shortgrass steppe was never under ice, but landscapes in the region currently demonstrate effects of the glacial/interglacial cycles; soils are commonly composed of mixtures of alluvium deposited as outwash from the melting of mountain glaciers. These cycles also had an enormous effect on the fauna. During the past 20,000 years, with the termination of the most recent glacial period, 32 genera of mammals have disappeared. These extinctions are thought by some researchers to have been the result of human hunters who entered North America from Siberia across the Bering Land Bridge during the Wisconsin period (Brown and Lomolino, 1998; Owen-Smith, 1987; Stuart, 1991). The Holocene (the past 11,000 years) has been characterized by significant climatic fluctuation in the northern shortgrass steppe. Muhs (1985) used data from dune fields in northeastern Colorado to reconstruct climate during the mid to late Holocene to suggest that the eolian sand was deposited during the Altithermal (from 8000 to 5000 years before present [BP]), a warm and dry period throughout much of western and central North America. The Altithermal was followed by

The Shortgrass Steppe: The Region and Research Sites 7

2000 years of cooler and wetter conditions, during which rates of soil formation exceeded rates of erosion. From 3000 to 1000 years BP, the climate again became drier, and erosion apparently exceeded soil formation. This second period of dune formation is when the modern dune fields in northeastern Colorado were created. Since 1500 years BP, conditions have been cooler and wetter, again allowing soil formation to exceed erosion, resulting in the development of modern soils and vegetation (Kelly, 1989; Kelly et al., chapter 3, this volume).

Geographic Features The shortgrass steppe covers approximately 3.4 × 105 km2 in the central and southern Great Plains, representing 11% of the central grassland region (Fig. 1.1) (Küchler, 1964; Lauenroth et al., 1999). The northern boundary between the shortgrass steppe and the northern mixed prairie is approximately the Colorado– Wyoming border, at 41ºN latitude, and coincides with the boundary between the Colorado Piedmont and the High Plains sections of the Great Plains. At this location, the High Plains is, on average, 200 m higher than the Colorado Piedmont. This elevation change affects the seasonality of both temperature and precipitation (Lauenroth and Milchunas, 1992). The northern boundary is slightly different from the northern boundary of Küchler’s Bouteloua–Buchloë type (Küchler, 1964; Lauenroth, chapter 5, this volume). The steppe extends southward to latitude 32ºN in western Texas, where it grades into southern mixed prairie to the east and Chihuahuan desert shrub savanna to the west (Küchler, 1964; Lauenroth et al., 1999). The western edge of the shortgrass steppe is the foothills of the Rocky Mountains, where it abuts coniferous woodland and forest. The key conifer species are Pinus ponderosa, Juniperus scopulorum, and Pseudotsuga menziesii. The eastern boundary of the shortgrass steppe with the southern mixed prairie occurs in western Kansas, western Oklahoma, and western Texas, reaching a maximum eastward extension of 100ºW longitude in Oklahoma.

Climatic Features The shortgrass steppe occurs in the driest and warmest portion of the Great Plains (Lauenroth and Burke, 1995; Pielke and Doesken, chapter 2, this volume). Mean annual precipitation ranges from 300 to 600 mm, with the lowest amounts occurring in the northwestern and southwestern portions, and the largest amounts along the eastern edge. Late spring to early summer is the wettest time of year; winter is the driest. Mean annual temperatures range from less than 9 ºC in the north to more than 16 ºC in the south. The northern portion of the region has an average of more than 170 days per year with daily temperatures reaching 0 ºC or less, whereas the southern portion has fewer than 70 such days. The shortgrass steppe lies entirely within the semiarid zone of the central grassland region as defined by Bailey (1979) (Lauenroth et al., 1999). The boundary between the shortgrass steppe and the southern mixed prairie is roughly the

8

Ecology of the Shortgrass Steppe

boundary between the semiarid and the dry subhumid zones in Kansas, Oklahoma, and Texas. The climate is discussed in detail by Pielke and Doesken in chapter 2 (this volume).

Vegetation Shortgrass steppe vegetation is characterized by the dominance of two C4 grasses: Bouteloua gracilis and Buchloë dactyloides. Both are perennial caespitose grasses, but B. dactyloides has the capability to produce stolons, giving it the potential to form a matlike stand and spread rapidly after disturbances (Aguiar and Lauenroth, 2001). Bouteloua gracilis is an archetypical caespitose grass that spreads by tillering. Approximately 70% of the shortgrass steppe remains in natural vegetation, and the majority of that is used for livestock grazing (Fig. 1.4). The rest of the area is used for row crops, divided between irrigated and dryland crops (Lauenroth et al., 1999). The major irrigated crops are Medicago spp. (alfalfa), Zea mays (corn), Beta vulgaris (sugar beet), and Gossypium spp. (cotton). Much of the irrigated land is adjacent to the major river channels or lies over the Ogallala aquifer. The dominant dryland crop is Triticum aestivum (wheat), which is grown using a summer fallow rotation system in which a crop is grown every other year. The land lies fallow during the alternate year, storing water for the crop year. The dominant grasses of the shortgrass steppe (B. gracilis and B. dactyloides) occur throughout the central grassland region, although B. dactyloides is less

Figure 1.4

Cattle on the shortgrass steppe. (Photo by Sallie Sprague.)

The Shortgrass Steppe: The Region and Research Sites 9

common in the northernmost part (Epstein et al., 1998). In many locations in the mixed prairies, heavy grazing by domestic livestock can result in a shift in dominance from the normally dominant midheight grasses to one or both of the short grasses. This has led to widespread use of the term shortgrass prairie (prairie is defined as having continuous plant cover whereas steppe is characterized by discontinuous cover) to describe grasslands from Canada to the Gulf of Mexico. Many of these areas are fundamentally different from the shortgrass steppe in that they have the potential to support midheight grasses with an understory of shortgrasses. The true shortgrass steppe lacks such potential on upland sites. The vegetation is described in detail by Lauenroth in chapter 5 (this volume).

Research Sites and Research Approaches The information contained in this book draws heavily on the results of long-term interdisciplinary studies conducted within one area encompassing the USDA ARS Central Plains Experimental Range (CPER) and the adjacent USDA Forest Service Pawnee National Grasslands (PNG); this area is designated as the Shortgrass Steppe Long-Term Ecological Research (SGS LTER) site (http://sgslter.colostate. edu) (Fig. 1.5). Two other locations have been sites of short-duration, concentrated research activity. The Texas Tech University Research Farm, 24 km east of Amarillo, Texas, was the focus of considerable research activity during the late 1960s and early 1970s, associated with the Grassland Biome project of the IBP. The other site, located near Springfield, Colorado, was a Colorado State University (CSU) Agricultural Experiment Station from 1957 to1998. The CPER encompasses 6280 ha of shortgrass steppe located approximately 8 km north of Nunn, Colorado (40º49⬘N latitude, 107º47⬘W longitude), and is

Wyoming

Nebraska

Utah

Arizona

Colorado

Kansas

New Mexico

Oklahoma Texas

Central Plains Experimental Range (CPER)

Wyoming

Nebraska

Colorado

Pawnee National Grassland

Figure 1.5 Location of the Shortgrass Steppe Long-Term Ecological Research site and the two administrative units within it: the Central Plains Experimental Range (CPER) and the Pawnee National Grasslands (PNG). Image courtesy of U.S. National Forest Service, Pawnee National Grasslands.

10

Ecology of the Shortgrass Steppe

operated by the Rangeland Resources Research Unit of the USDA–ARS. Average elevation is 1650 m, mean annual precipitation is 321 mm, and mean annual temperature is 8.6 ºC. The majority of annual precipitation falls as rain during the May to September growing season. Mean monthly temperatures range from–5 ºC in January to 22 ºC in July. The CPER was established in 1939 on land taken over by the USDA Forest Service after drought, overgrazing, and dust storms had taken their toll on ranches and farms in eastern Colorado, forcing their abandonment (Shoop et al., 1989). Forest Service scientists began a research program directed toward developing sustainable management practices, with the particular objective of enhancing livestock production from shortgrass rangelands. From the beginning, all livestock for research were contributed by the Crow Valley Livestock Cooperative, Inc., composed today of approximately 40 individual ranches. As a result, much of the station’s research has been conducted in close collaboration and with the strong support of the local ranching community. The first formal research began in May 1939, with a grazing intensity study that is still underway today. In 1953, the Agricultural Act transferred the CPER to the USDA–ARS. Early research at the CPER determined critical relationships among stocking rate, forage production, and animal gain that provided foundational information for establishing recommended grazing practices for the region (Bement, 1969; Hyder et al., 1975; Klipple and Costello, 1960). Research by the ARS in the 1960s and 1970s focused on reseeding, legume genetics, range improvements, and plant–grazing interactions (Shoop et al., 1989; Townsend et al., 1995). Many of the results used in this book come either directly from publications by ARS scientists or from experiments that they initiated. Researchers from CSU and from other nonfederal organizations had a relatively small presence at the CPER until the late 1960s, when the National Science Foundation (NSF) initiated funding of the U.S. IBP Grassland Biome project. The Grassland Biome project was led by the visionary scientist Dr. George Van Dyne, and for a period of 9 years it received approximately $16 million, which funded the work of some 200 scientists (Golley, 1993; Van Dyne and Anway, 1976). A large proportion of that funding was spent on research at the CPER. Van Dyne’s vision was that the Grassland Biome project be the first research project to take a systems approach to grasslands, including a major role for simulation modeling (Van Dyne and Anway, 1976). Although the Grassland Biome project was not successful in achieving all of its objectives, in retrospect it was one of the most important building blocks of modern ecosystem science. Despite the fact that he was naive about many of the details, Van Dyne’s belief in the importance of a systems approach to studying ecology has been upheld since he first articulated it. Much of the data included in this volume were collected during the Grassland Biome project. Funding for the IBP formally ended in 1974, and from then until 1982 research on the CPER was continued by ARS and CSU scientists who were funded by individual research grants. In 1982, the CPER was funded as an NSF LTER site (Franklin et al., 1990). The overall objective of the SGS LTER project has been to understand the processes that account for the origin and maintenance of the

The Shortgrass Steppe: The Region and Research Sites 11

structure and function of shortgrass steppe ecosystems. During the first 15 years of the project (1982–1996), effectively all the research was conducted on the CPER. After 1996, the research area was expanded to include portions of the PNG to incorporate more of the region’s variability in soils, vegetation, and land management (Burke and Lauenroth, 1993). Initially, the SGS LTER was under the direction of Drs. Robert Woodmansee and William Lauenroth, both CSU scientists. Both Woodmansee and Lauenroth were trained under Van Dyne’s influence during the IBP project, Lauenroth as a PhD student and Woodmansee as a postdoctoral fellow. Many of the participants of the newly funded LTER project had been members of the IBP team. The original leadership also included Dr. William Laycock, Research Leader of the Forage and Range Research Unit of the ARS. Thus, the SGS LTER integrated the longterm experience and experiments of the ARS with those of the IBP. The ARS range program began shifting in the 1980s more toward sustainability, conservation, and environmental concerns, and continues in that direction today. However, current research still addresses the early interests in beef cattle production (Derner and Hart, 2005; Hart and Ashby, 1998), shrub ecology (Cibilis et al., 2003a, b), and range improvements (Shoop et al., 1985, 1989). Recent initiatives in trace gases (Mosier et al., 1991, 1996), climate change (Morgan et al., 2004, 2007), and plant–animal interactions (Derner et al., 2006) have helped forge a closer vision between ARS and CSU scientists, and have enhanced their ability to conceptualize and implement relevant ecological rangeland research for a diverse set of clients. The SGS LTER project has made a number of major contributions in the area of grazing ecology. The CPER, with its long-term grazing experiments initiated in the 1930s and maintained since then by the ARS, has been a key ingredient to the group’s success. Perhaps more important than the actual field experiments and facilities has been the gathering of agriculturalists, ecologists, modelers, and many others with an interest in native grasslands at a single site where critical and honest discourse are encouraged. This combination of resources has been instrumental in advancing a systems approach to grassland science. This systems approach, which has guided the SGS LTER project since its inception in 1982, contributed substantially to the training of the first principal investigators of the SGS LTER project and is a direct result of the influence of George Van Dyne and the Grassland Biome project. In this book, shortgrass steppe researchers describe the history, environment, ecology, and vulnerabilities of the shortgrass steppe as we understand them today. Our hope is that future generations will find our analyses to be useful steppingstones from which to continue research into the mysteries of the shortgrass steppe.

References Axelrod, D. I. 1985. Rise of the grassland biome. Botanical Review 51:163–201. Aguiar, M. R., and W. K. Lauenroth. 2001. Local and regional differences in abundance of co-dominant grasses in the shortgrass steppe: a modeling analysis of potential causes. Plant Ecology 156:161–171.

12

Ecology of the Shortgrass Steppe

Bailey, H. P. 1979. Semi-arid climates: Their definition and distribution, pp. 73–97. In: A. E. Hall, G. H. Cannell, and H. W. Lawton (eds.), Agriculture in semi-arid environments. Springer-Verlag, Berlin. Bement, R. R. 1969. A stocking-rate guide for beef production on blue-grama range. Journal of Range Management 22:83–86. Burke, I. C., and W. K. Lauenroth. 1993. What do LTER results mean? Extrapolating from site to region and decade to century. Ecological Modelling 67:19–35. Brown, J. H., and M. V. Lomolino. 1998. Biogeography. Sinauer Associates, Inc., Sunderland. Cibils, A. F., D. M. Swift, and R. H. Hart. 2003a. Changes in shrub fecundity in fourwing saltbush browsed by cattle. Journal of Range Management 58:39–46. Cibils, A. F., D. M. Swift, and R. H. Hart. 2003b. Female-biased herbivory in fourwing saltbush browsed by cattle. Journal of Range Management 56:47–51. Cushman, R. C., and S. R. Jones. 1988. The shortgrass prairie. Pruett Publishing, Boulder, Colo. Derner, J. D., J. K. Detling, and M. F. Antolin. 2006. Are livestock gains affected by blacktailed prairie dogs? Frontiers in Ecology and Environment 4:459–464. Derner, J. D., and R. H. Hart. 2005. Heifer performance under two stocking rates on fourwing saltbush-dominated rangeland. Rangeland Ecology and Management 58:489–494. Ehleringer, J. R., T. E. Cerling, and B. R. Helliker. 1997. C4 photosynthesis, atmospheric CO2, and climate. Oecologia 112:285–299. Epstein, H. E., W. K. Lauenroth, I. C. Burke, and D. P. Coffin. 1998. Regional productivities of plant species in the Great Plains of the United States. Plant Ecology 134:173–195. Franklin, J. F., C. S. Bledsoe, and J. T. Callahan. 1990. Contributions of the Long-Term Ecological Research program. BioScience 40:509–523. Golley, F. B. 1993. A history of the ecosystem concept in ecology. Yale University Press, New Haven, Conn. Hart, R. H., and M. M. Ashby. 1998. Grazing intensities, vegetation, and heifer gains: 55 years on shortgrass. Journal of Range Management 51:392–398. Hyder, D. N., R. E. Bement, E. E. Remmenga, and D. F. Hervey. 1975. Ecological responses of native plants and guidelines for management of shortgrass range. USDA technical bulletin no. 1503. USDA. Washington, D.C. Kelly, E. F. 1989. A study of the influence of climate and vegetation on the stable isotope chemistry of soils in grassland ecosystems of the Great Plains. PhD diss., University of California, Berkeley, Calif. Klipple, G. E., and D. F. Costello. 1960. Vegetation and cattle responses to different intensities of grazing on shortgrass ranges on the Central Great Plains. USDA technical bulletin no. 1216. USDA. Washington, D.C. Küchler, A. W. 1964. Potential natural vegetation of the conterminous United States. American Geographical Society, New York. Lauenroth, W. K., and I. C. Burke. 1995. Great Plains: Climate variability, pp. 237–249. In: W. A. Nierenberg (ed.), Encyclopedia of environmental biology. Academic Press, New York. Lauenroth, W. K., I. C. Burke, and M. P. Gutmann. 1999. The structure and function of ecosystems in the central North American grassland region. Great Plains Research 9:223–259. Lauenroth, W. K., and D. G. Milchunas. 1992. Short-Grass Steppe, pp. 183–226. In: R. T. Coupland (ed.), Natural grasslands: Introduction and western hemisphere. Ecosystems of the world, vol. 8A. Elsevier, Amsterdam.

The Shortgrass Steppe: The Region and Research Sites 13 Michener, J. A. 1974. Centennial. Fawcett Publishing, Greenwich, Conn. Morgan, J. A., D. G. Milchunas, D. R. LeCain, M. West, and A. Mosier. 2007. Carbon dioxide enrichment alters plant community structure and accelerates shrub growth in the shortgrass steppe. Proceedings of the National Academy of Sciences 104:14724–14729. Morgan, J. A., A. R. Mosier, D. G. Milchunas, D. R. LeCain, J. A. Nelson, and W. J. Parton. 2004. CO2 enhances productivity, alters species composition, and reduces forage digestibility of shortgrass steppe vegetation. Ecological Applications 14:208–219. Mosier, A. R., W. J. Parton, D. W. Valentine, D. S. Ojima, D. S. Schimel, D. J. Delgado. 1996. CH4 and N2O fluxes in the Colorado Shortgrass Steppe: I. Impact of landscape and nitrogen addition. Global Biogeochemical Cycles 10:387–399. Mosier, A. R., D. S. Schimel, D. Valentine, K. Bronson, and W. J. Parton. 1991. Methane and nitrous oxide fluxes in native, fertilized and cultivated grasslands. Nature 350:330–332. Muhs, D. R. 1985. Age and paleoclimatic significance of Holocene sand dunes in northeastern Colorado. Annals of the Association of American Geographers 75:566–582. NLCD. 2001. National Land Cover Database. Online. Available at http://www.mrlc.gov/ mrlc2k_nlcd.asp. Owen-Smith, N. 1987. Pleistocene extinctions: The pivotal role of megaherbivores. Paleobiology 13:351–362. Shoop, M. C., R. C. Clark, W. A. Laycock, and R. J. Hansen. 1985. Cattle diets on shortgrass ranges with different amount of fourwing saltbush. Journal of Range Management 38:443–449. Shoop, M., S. Kanode, and M. Calvert. 1989. Central Plains Experimental Range: 50 years of research. Rangelands 11:112–117. Stebbins, G. L. 1981. Coevolution of grasses and herbivores. Annals of the Missouri Botanical Garden 68:75–86. Stuart, A. J. 1991. Mammalian extinctions in the late Pleistocene of northern Eurasia and North America. Biological Review of the Cambridge Philosophical Society 66:453–562. Townsend, C. E., S. Wand, and T. Tsuchiya. 1995. Registration C-25, C-26, and C-27 alfalfa germplasms. Crop Science 35:289. Van Dyne, G. M., and J. C. Anway. 1976. A program for and the process of building and testing grassland ecosystem models. Journal of Range Management 29:114–122.

2 Climate of the Shortgrass Steppe Roger A. Pielke, Sr. Nolan J. Doesken

T

he climate of a region involves the short- and long-term interaction among the atmospheric, hydrologic, ecologic, oceanographic, and cryospheric components of the earth’s environmental system (Hayden, 1998; Pielke, 1998, 2001a,b). These interactions occur across all spatial and temporal scales, from turbulence generated by diurnal cycles at a landscape scale, to globalscale circulation. The establishment of particular ecosystem types is associated with a nonlinear feedback between the atmosphere and the underlying vegetation (Pielke and Vidale, 1995). Wang and Eltahir (2000) and Claussen (1998) have demonstrated that vegetation patterning cannot be accurately simulated in a model unless vegetation–atmosphere feedbacks are included. In this chapter we summarize the climate system of the shortgrass steppe. This is a region of large seasonal contrasts, and of interannual and longer term variability. It is also a region that has undergone major human impacts during the past 150 years. We present both average conditions and examples of extreme events in the shortgrass steppe to illustrate the variable climate of this interesting ecosystem.

Geographic Factors Controlling the Climate Geographic factors play a large role in determining the climatic characteristics of the shortgrass steppe (Lauenroth and Burke, 1995; Lauenroth and Milchunas, 1992; Lauenroth et al., 1999). Key factors for this region include its mid-latitude position, its relatively high elevations, its interior continental location, and its 14

Climate of the Shortgrass Steppe 15

proximity to the Rocky Mountains, a substantial north–south-oriented mountain barrier immediately to the west. Air masses affecting the region consist of continental polar air from the north, humid continental air masses from the east, humid subtropical air masses from the southeast and south, and Pacific maritime air masses from the west. The latter can be significantly modified as they cross a series of mountain ranges and interior dry regions before reaching the shortgrass steppe region. Each of these geographic and atmospheric features contributes to the climate of the region. Latitude determines day length and sun angle, and, hence, solar insolation. This, in turn, greatly affects air temperature. Upper level westerly winds increase over the mid-latitudes in the fall and winter in response to strengthening north–south temperature gradients in the atmosphere. Pacific air masses are carried eastward over the Rocky Mountains, depositing considerable cool-season precipitation in the mountains, but rarely on the shortgrass steppe. The high elevation of the region contributes to its low humidity and intense solar insolation. This also means that infrared radiational heat losses from the earth to space at night lead to rapid cooling of the air near the ground, resulting in large diurnal temperature variations. The interior continental location of the region contributes further both to large day-to-day and large seasonal temperature changes. Atmospheric moisture must travel long distances to reach the area, unlike the coastal regions of the continent. The mountain ranges to the west play a huge role in the region’s climate by serving as an effective block or rain shadow from storms carrying Pacific moisture from the west. The mountains impose both upward motion on the windward side and downward vertical motion on the leeward side. This is critical to cloud formation and precipitation. Throughout the winter season, westerly winds are dominant aloft over the mid-latitudes, bringing predominant downslope winds to the shortgrass steppe as the air descends the eastern slopes of the Rocky Mountains. This translates into abundant sunshine, low humidity, low precipitation, and periods of strong winds in the shortgrass steppe. But occasionally, for brief periods of time ranging from a few hours to a few days, easterly upslope winds ascend the High Plains, bringing low clouds and widespread precipitation to the region. This occurs in the fall and winter, but most frequently in March, April, and May, when slow-moving storms are able to tap into moisture from the Gulf of Mexico and carry it into the region from the southeast. During summer, winds aloft become weak. Air masses advect more slowly, and local factors become more important. Convection becomes the primary mechanism for producing upward motion, which leads to cloud formation and precipitation. Dry air from the intermountain area of the southwestern United States is the most common warm-season air mass in the region, but much more humid air lies just east and southeast of the shortgrass steppe. Under certain weather patterns, this moist air is pushed westward, fueling occasional heavy, local storms. During late summer, the North American monsoon, which is strongly influenced by the Mexican Plateau and elevated terrain in the southwest United States (Castro et al., 2001), provides an additional source of moisture, particularly over the southern portion of the area.

16

Ecology of the Shortgrass Steppe

Overview of the Weather Patterns in the Shortgrass Steppe The geographic factors described earlier work together to produce a climate characterized by a strong annual cycle, large daily variations, frequent and persisting dry weather, and occasional very vigorous storms. Winter The winter weather over the shortgrass steppe is dominated by frequent migratory high- and low-pressure systems that are associated with the polar jet stream and its associated polar front. When the jet stream flow is zonal, air masses associated with the front are Pacific in origin, producing substantial snows in the mountains and relatively mild air in the shortgrass steppe just to the lee of these barriers. Further east, the weather in the Great Plains is cooler, with some air from Canada entrained south into the region west of the low-pressure systems. Occasionally, Arctic high-pressure systems travel southward over the region, producing the area’s coldest weather. These intrusions of Arctic air occur when the polar jet stream travels across Alaska and northwest Canada, before plunging south over the Great Plains. Upslope snows frequently occur in the western High Plains during these cold outbreaks. The shortgrass steppe coincides with a preferred region of strong cool-season cyclogenesis (formation zone for low-pressure areas) (Davis et al., 1999). Very strong winds (Weaver, 1999) and rapidly changing weather conditions occur as these storm systems develop and sweep across the Plains and Midwest. Spring Spring is a transition season when high sun angles can produce warm days, yet cold air masses still occasionally travel southward bringing heavy snows to the Plains. Extreme weather changes are common in spring as cold and warm air masses alternate. In the western High Plains, March and April are the snowiest months of the year. Ferocious blizzards are not uncommon during the spring— decreasing from south to north as the sun climbs higher in the sky. Increased solar heating of the elevated land surfaces rapidly heats the ground while temperatures aloft remain cool. This thermal instability effectively mixes the atmosphere, helping bring strong winds down to the surface. The result is many very windy days during spring. At this time in the eastern High Plains, thunderstorms become common. Thunderstorms can often be quite intense during spring as a result of strong solar surface insolation, relatively cold air aloft, and sharp contrasts in air masses. The tornado season in the shortgrass steppe region peaks during April through June as a still-vigorous polar jet stream provides large changes in wind speed and direction with altitude. This large wind shear provides the initial horizontal wind circulation for tornadic thunderstorms, particularly when the wind shear is tilted on its side by intense thunderstorm updrafts and downdrafts. A tornado can subsequently be produced when this horizontal wind circulation is concentrated into a small area by intense updrafts.

Climate of the Shortgrass Steppe 17

However, spring is also characterized by slower migration of large, horizontallength waves in the atmosphere that control the motion of surface lows and highs. This results in slow-moving storms that have time to tap into abundant humidity east of the region. Spring storms bring occasional episodes of precipitation (either rain or snow) that may last for 1 to 3 days and cover broad areas. This moisture, falling just as the area is greening up in the spring, is important to local vegetation. Summer By summer, the polar jet has typically migrated far to the north. Rainfall becomes dominated by topographically heated upslope flows, weak migratory low-pressure systems, local air mass boundaries, and storm outflows. During this period, weather pattern changes are relatively slow. Precipitation is characterized by thunderstorms that bring high-intensity rainfall, but over relatively small areas for short periods of time. As storms progress eastward and tap into more humid air, they produce more widespread rainfall. Organized clusters of thunderstorms, called mesoscale convective systems (MCSs), can develop and are associated with weak frontal boundaries or higher terrain (Cotton, 1999). These MCSs usually move eastward in response to the weak westerly winds in the middle and upper troposphere during this time of the year. A dryline boundary usually forms in the western Plains, separating humid air coming from the Gulf of Mexico from dry air originating in the desert Southwest and northern Mexico (Ziegler, 1999). During summer, thunderstorms frequently form along this dryline boundary. In late summer, the North American monsoon starts to affect the western portion of this region. Substantial rains often occur as moisture originating in the tropical Pacific Ocean or the southern Gulf of Mexico is advected northward into the southern Plains and southern Rocky Mountains. Autumn The North American monsoon flow regime weakens in late August, and relatively dry weather, dominated by persistent and often nearly stationary high-pressure systems, begins to dominate fall months. Long stretches of fair weather are interrupted occasionally by cool, cloudy, and wet periods as mid-latitude storm systems develop in response to the changing seasons. Sometimes, widespread heavy rains occur as moisture from tropical storms is entrained. Snowstorms can occur as early as September in the northern and central Great Plains and Rocky Mountains, when the polar jet stream migrates southward in early fall. Summary The climate of the shortgrass steppe displays a strong seasonal cycle. Each season is characterized by significantly different meteorological conditions that are described by weather events such as the passage of frontal systems, development of leeside low-pressure troughs and low-pressure centers, and the production of

18

Ecology of the Shortgrass Steppe

convective storms. These weather events operate on timescales ranging from several hours to a few days, resulting in exciting and changeable weather conditions.

Temperature The shortgrass steppe is an area of large and rapid temperature changes. A single year’s daily temperature data for the SGS LTER site provides a clear illustration of the large diurnal and seasonal temperature changes that are characteristic of the region (Fig. 2.1). The shortgrass steppe region is also characterized by temperature gradients from west to east and from north to south (Fig. 2.2). Hot temperatures are common across the area during summer, especially at low elevations and over the easternmost portions of the region. Record maxima in this region have exceeded 42 ºC (Fig. 2.3). There is a negative correlation between precipitation and temperature. When summer precipitation has been significantly lower than average, summer temperatures are especially hot. Under these circumstances, more of the solar insolation is used in heating the ground and adjacent air whereas less is utilized in evaporating water (Pielke, 2001b). The shortgrass steppe is also characterized by periods of cold temperatures in winter. The coldest weather occurs in the northern and highest elevation portion of the steppe. During an average year, along the northern edge of the shortgrass steppe, temperatures occasionally fall to around –34 ºC. In extreme years,

40 30

Temperature (C)

20 10 0 Freezing

-10 -20 -30 0

30 60 90 120 150 180 210 240 270 300 330 360 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec

Figure 2.1 Daily maximum and minimum temperatures for an entire year from the Shortgrass Steppe LTER site.

Climate of the Shortgrass Steppe 19 (A)

(B)

Legend (January maxima)

Legend (January minima)

0.1 - 4.4 C 4.5 - 10.0 C 10.1 - 15.5 C

(C)

-17.7 to -9.4 C -9.3 to -3.9 C -3.8 to 0.0 C

(D)

Legend (July maxima) 21.2 - 26.6 C 26.7 - 32.2 C 32.3 - 37.8 C

Legend (July minima) 10.1 - 15.5 C 15.6 - 21.1 C

Figure 2.2 (A) January mean maximum temperatures. (B) January mean minimum temperatures. (C) July mean maximum temperatures. (D) July mean minimum temperatures for the shortgrass steppe. (Adapted from NCDC [2002].)

temperatures less than –40 ºC can occur (Fig. 2.3). Even in the southern portion of the shortgrass steppe, temperatures can fall to less than –12 ºC in extreme years. The large range between summer and winter temperatures, which can exceed 49 ºC in the northern portion of the steppe (Fig. 2.2), influence which animals and plants can survive in this region. Growing season characteristics are another indicator of a region’s climate. Vegetation, especially many agricultural crops, can be sensitive to the first and last occurrence of 0 ºC during the year (Fig. 2.4). The entire shortgrass steppe experiences periods of the year with subfreezing temperatures ranging from just

20

Ecology of the Shortgrass Steppe

(A)

(B)

Legend (extreme maxima) 32.3-35.0 C 35.1-37.8 C 38.3-40.5 C 41.1-43.3 C 43.9-46.1 C

Legend (extreme minima) < -40.0 C -40.0 - -34.4 C -34.3 - -28.9 C -28.8 - -23.3 C -23.2 - -17.7 C

Figure 2.3 (A) Annual record extreme maximum temperatures. (B) Annual record extreme minimum temperatures for the shortgrass steppe. (Adapted from NCDC [2002].)

more than 200 days in southeast Wyoming to fewer than 90 days in western Texas and southeastern New Mexico (NCDC, 2002). In addition, the region experiences many freeze–thaw cycles through winter as temperatures warm during the day but cool quickly at night. The frequent freezing and thawing of the surface topsoil plays an interesting and important role in soil conditioning, potentially making bare soil very vulnerable to wind erosion during late winter and spring (Doesken, 1988).

Humidity Humidity over the shortgrass steppe is characteristically low much of the year, particularly when the region is dominated by air masses crossing the mountains from the west or moving down the continent from the north. However, high-humidity air often lies just east and southeast of the shortgrass steppe and does move westward under appropriate meteorological conditions (NCDC, 2002). Relative humidity is a common measure of atmospheric moisture; however, it can be misleading because it is very much an inverse function of air temperature. The dew point temperature is a more appropriate measure of humidity, because it is the temperature at which condensation would begin to occur if the air is cooled without changing its pressure or water content. Thus, because it is an absolute measure of the amount of water vapor in the air, it is the index of atmospheric moisture most often used by meteorologists for tracking and comparing humidity. The annual average maximum dew point temperatures decrease greatly from southeast to northwest across the shortgrass steppe (Fig. 2.5) as elevation, latitude, and distance from the Gulf of Mexico moisture increases.

Climate of the Shortgrass Steppe 21 (A)

(B)

Legend (median last freeze) Mar 1 - Mar 31 Apr 1 - Apr 15 Apr 16 - Apr 30 May 1 - May 15 May 16 - May 31

Jun 1 - Jun 30

Legend (median first freeze) Sep 1 - Sep 30 Oct 1 - Oct 15 Oct 16 - Oct 31 Nov 1 - Nov 15

(C)

Legend (frost free days) 91 - 120 121 - 180 181 - 240

Figure 2.4 (A) Median date of last freeze in spring. (B) First freeze in fall. (C) Length of frost-free period for the shortgrass steppe. (Adapted from NCDC [2002].)

Precipitation Precipitation is arguably the single most important climatic variable controlling the ecology of the shortgrass steppe (Lauenroth and Sala, 1992). Lu et al. (2001) showed that the response of vegetation to precipitation in the central U.S. grasslands is the area’s largest weather sensitivity. This followed up on work by Sala et al. (1988) and Lauenroth and Sala (1992), showing that precipitation is the single largest predictor of shortgrass steppe productivity both across regions and among years. Pielke et al. (1999) have shown that soil moisture at the beginning of the growing season, resulting from winter and fall precipitation, influences subsequent spring and summer rainfall in the shortgrass steppe, with moister soils

22

Ecology of the Shortgrass Steppe

Legend (annual mean maximum dewpoint) < - 1.1 C 1.1 - 1.7 C 1.8 - 4.4 C 7.3 - 10 C

Figure 2.5 Annual mean maximum dew point temperatures for the shortgrass steppe. (Adapted from NCDC [2002].)

favoring enhanced rainfall. This positive feedback occurs as enhanced evaporation and transpiration from the wetter soils provide additional energy to fuel more rainfall-producing thunderstorms. Eastman et al. (2001a,b) have shown how settlement and changes in land use, such as the removal of bison and the switch from livestock to crop production, has altered precipitation in the region. Mean annual precipitation of the shortgrass steppe varies from 508 to 635 mm over the easternmost portions to less than 305 mm in some locations farther west (Fig. 2.6). Precipitation is highly seasonal. Growing season (April–September) precipitation contributes from 70% to 82% of the average annual moisture across the region (NCDC, 2002). Unlike temperature, humidity, and solar radiation, which are continuous weather variables, precipitation is episodic. Precipitation only falls in 2% to 4% of the hours of the year in the shortgrass steppe (Doesken and Eckrich, 1987). As in many semiarid regions, a few storms during the year produce a large portion of the annual precipitation (Sala et al., 1992). Cowie and McKee (1986) showed that in eastern Colorado, 20% of the days with measurable precipitation account for at least 50% of the accumulated precipitation. At the SGS LTER site, less than 10% of the precipitation events produce more than 20 mm of water, but these large events contribute more than 30% of annual precipitation (Sala and Lauenroth, 1982). This pattern extends throughout most of the shortgrass steppe region (Fig. 2.7). Very intense, potentially erosive, but fairly localized rainfall is a common trait of the shortgrass steppe (Hjelmfelt, 1999). Storm rainfall that occurs for a 6-hour duration only has a 1% probability of occurrence at a specific point in any given year. The size of these events ranges from about 76 mm over the northern portion of the region up to more than 127 mm over the southeastern portion of the region. Storms equal to or greater than this occur several times almost every year somewhere in the area (NCDC, 2002). Vegetation becomes very important in minimizing the water erosion associated with these heavy downpours. Doesken

Climate of the Shortgrass Steppe 23

Legend (annual mean total precipitation) 127 - 305 mm 305.1 - 508 mm 508.1 - 762 mm

Proportion of total events

Figure 2.6 Annual mean total precipitation for the shortgrass steppe. (Adapted from NCDC [2002].)

0.80 Weld, CO Kit Carson, CO Finney, KS Texas, OK Quay, NM Midland, TX

0.70 0.60 0.50 0.40 0.30 0.20 0.10 0.00 0 - 100 m) that could account for recovery patterns on old fields (Fraleigh, 1999). Similar studies of dispersal vectors for B. dactyloides found that dispersal by wind is limited to the immediate area around a plant, but that dispersal by cattle feces and hair may provide long-distance movement of seeds (Fraleigh, 1999). Another poorly understood aspect of old-field dynamics in the shortgrass steppe is the long-term recovery of soil organic matter and soil fertility after abandonment. Our companion studies of carbon and nitrogen dynamics conducted on the same 13 fields have provided important information on soil processes and properties. Our analyses compared native (never cultivated), abandoned (cultivated

The Role of Disturbances in Community and Ecosystem Dynamics 111

until 1937), and currently cultivated fallow fields. Our results show that 50 years after abandonment, surface (0–10 cm) and subsurface soils (0–30 cm) are 30% to 40% lower in carbon and nitrogen compared with nearby native fields (Burke et al., 1995; Ihori et al., 1995b). Microbial biomass and nitrogen mineralization, as indexed by potential net carbon and nitrogen mineralization, are also significantly reduced (Burke et al., 1995). By contrast, rates of nitrogen mineralization and turnover are highest in cultivated fields, likely as a result of higher soil water content (Ihori et al., 1995a). Microbial biomass, potentially mineralizable nitrogen, and respirable carbon are not significantly different between abandoned and native fields (Burke et al., 1995). Small-scale heterogeneity in soil carbon and nitrogen associated with individual B. gracilis plants also recovered on abandoned fields within the 53-year recovery period. Measured variation in soil properties is less on abandoned fields than variation in vegetation, because sampling was only conducted under B. gracilis plants. High spatial variation within and among fields would likely have been found if sampling sites had not been limited to this one species. For example, we found that annuals do not accumulate significant levels of nutrients in soils beneath their canopies (Vinton and Burke, 1995). Our results also show that recovery of vegetation and soil properties on old fields is highly variable, and depends upon a number of historical factors as well as more recent climatic conditions. The number of years and intensity of cultivation, subsequent grazing intensity, soil texture, proximity to native seed sources, slope and topography, kind and degree of erosion, amount of soil deposition, amount and distribution of rainfall, and distance, direction, and speed of wind have been identified as potentially important to recovery rates (Costello, 1944; Judd, 1974). Most abandoned fields in the region were reseeded at some time between 1940 and 1955 with a mixture of native and introduced species, including Agropyron cristatum. Differential success of these reseeding efforts on different fields may also have influenced the recovery rate of native grasses. We have very little understanding of how these various factors affect recovery rates, in large part because of the lack of historical information about each field. Drought Periodic drought occurs in the central Great Plains, including eastern Colorado (Pielke and Doesken, chapter 2, this volume). Low precipitation, high temperatures, high winds, and dust storms combine to kill plants as a result of severe water limitation and burial of plants by loose soil (Albertson and Weaver, 1942, 1944a). Two of the most important droughts during the past century (1933–1939 and 1952–1955) have been studied in terms of plant death and vegetation recovery. In eastern Colorado, effects of drought were more severe in the 1950s than in the 1930s. In the 1950s, east–central Colorado (average losses in cover of 89%) was more severely affected than northeastern Colorado (losses of 49% to 82%) (Albertson et al., 1957). During both droughts, B. gracilis was more tolerant of extreme conditions than B. dactyloides throughout eastern Colorado and western Kansas (Albertson and Weaver, 1944b; Albertson et al., 1957). Sporobolus cryptandrus, through abundant seedling establishment, also survived the droughts.

112

Ecology of the Shortgrass Steppe

Rate of recovery of dominant grasses after drought is highly variable and is related to a number of factors, including plant composition of predrought communities, intensity of drought, degree of plant survival, grazing intensity, and amount and distribution of precipitation after the cessation of drought (Albertson and Weaver, 1944b). In general, rate of recovery is much faster for B. dactyloides than for B. gracilis. In one pasture, B. dactyloides increased from 6.1% to 78.2% whereas B. gracilis only doubled its basal cover (7.3% to 15.1%) during the same time period (1940–1943). Our more recent studies have found that B. gracilis can respond rapidly to small rainfall events, and the ability of this species to respond after drought is related to the intensity and length of the drought (Sala and Lauenroth, 1982; Sala et al., 1981). The stages of succession after drought are similar to those for other disturbance types and consist of (1) annual forbs and grasses, including S. kali, Chenopodium spp., Plantago spp., H. annuus, and V. octoflora; (2) short-lived perennial grasses dominated by S. cryptandrus, A. smithii, S. paniculatus, and M. squarrosa; and (3) long-lived perennial grasses that included A. longiseta, S. hystrix, B. dactyloides, and B. gracilis (Albertson and Weaver, 1944b). Although actual rates of recovery were not determined, it was estimated that cover of dominant grasses could be restored within 5 years under wet periods with reasonable grazing management, but that more than 20 years would be required for pastures that were overgrazed prior to the drought (Albertson et al., 1957). Faster recovery compared with abandoned agricultural fields likely results from the survival of some small plants and tillers that could respond after the cessation of drought.

Summary and Conclusions The disturbance regime of shortgrass ecosystems consists of a rich array of disturbance types, each with its unique set of characteristics having important effects on individual plants and communities. Recovery patterns and rates after disturbance are dependent upon interactions among disturbance characteristics and the life history traits of plants. One of the most important characteristics is the size of the disturbed area. In general, patterns of recovery are similar among different sizes of disturbances, in that annuals and short-lived perennials colonize initially, with long-lived perennials eventually dominating the community. However, our results show that the rate of recovery is affected by disturbance size. Recovery rates increase in a nonlinear way as size increases. Disturbance intensity also modifies recovery patterns and rates. Low-intensity disturbances, such as areas killed by grubs or covered with soils by burrowing animals, do not completely kill perennial plants or their belowground organs, and thus have faster rates of recovery compared with high-intensity disturbances, such as old fields, where all plants are killed. Although grazing by cattle, soil texture, and topographic position have important effects on disturbance frequencies and sizes, very little is known about how these factors affect plant recovery. Our gap dynamics conceptualization of plant communities provides important insights into the role of disturbance in shortgrass ecosystems and is a better

The Role of Disturbances in Community and Ecosystem Dynamics 113

representation of shortgrass steppe dynamics than either the traditional Clementsian model or the conventional model modified for eastern Colorado. Actual recovery rates of B. gracilis on disturbances of different sizes can only be explained through our gap dynamics approach that explicitly includes scale-dependent processes interacting with disturbance characteristics. High variation in B. gracilis recovery between old fields with similar abandonment date and soil texture indicates the importance of historical factors, such as grazing management, reseeding practices, and local variations in precipitation and temperature, to recovery. Very fast rates of recovery by B. gracilis on some fields (>90 m within 52 years) suggest that biotic factors, such as dissemination of seed by cattle and ants, as well as abiotic events, such as extreme wind gusts, may also be important in the recovery of this species. Because of the importance of B. gracilis to shortgrass steppe ecosystems, and the controversy concerning its recovery after disturbance, much of the process-level research has focused on this species. Little is known about key processes limiting the response of other species to disturbance. Further research is needed to elucidate this information, which could be important in the management of these ecosystems, especially in terms of preserving biodiversity at multiple spatial and temporal scales.

Acknowledgments This research was supported by an NSF grant (BSR 9011659) to CSU as part of the SGS LTER program. We thank Brandon Bestelmeyer for assistance with the figures, and Tom Crist and other anonymous reviewers for helpful comments on the manuscript.

References Aguilera, M. O. 1992. Intraspecific interactions in blue grama. PhD diss., Colorado State University, Fort Collins, Colo. Aguilera, M. O., and W. K. Lauenroth. 1993a. Neighborhood interactions in a natural population of the perennial bunchgrass Bouteloua gracilis. Oecologia 94:595–602. Aguilera, M. O., and W. K. Lauenroth. 1993b. Seedling establishment constraints in adult neighborhoods: Intraspecific constraints in the regeneration of the bunchgrass Bouteloua gracilis. Journal of Ecology 81:253–261. Aguilera, M. O., and W. K. Lauenroth. 1995. Influence of gap disturbances and type of microsites on seedling establishment in Bouteloua gracilis. Journal of Ecology 83:87–97. Albertson, F. W., G. W. Tomanek, and A. Riegel. 1957. Ecology of drought cycles and grazing intensity in grasslands of central Great Plains. Ecological Monographs 27:27–44. Albertson, F. W., and J. E. Weaver. 1942. History of the native vegetation of western Kansas during seven years of continuous drought. Ecological Monographs 12:23–51. Albertson, F. W., and J. E. Weaver. 1944a. Effects of drought, dust, and intensity of grazing on cover and yield of short-grass pastures. Ecological Monographs 14:1–29. Albertson, F. W., and J. E. Weaver. 1944b. Nature and degree of recovery of grassland from the Great Drought of 1933 to 1940. Ecological Monographs 14:394–479.

114

Ecology of the Shortgrass Steppe

Anonymous. 1959. Control of common white grubs. USDA Farmers bulletin no. 1798. U. S. Department of Agriculture, Washington, D.C. Archer, S., and J. K. Detling. 1986. Evaluation of potential herbivore mediation of plant water status in a North American mixed-grass prairie. Oikos 47:287–291. Archer, S., M. G. Garrett, and J. K. Detling. 1987. Rates of vegetation change associated with prairie dog (Cynomys ludovicianus) grazing in North American mixed-grass prairie. Vegetatio 72:159–166. Bonham, C. D., and A. Lerwick. 1976. Vegetation changes induced by prairie dogs on shortgrass range. Journal of Range Management 29:221–225. Bormann, F. H., and G. E. Likens. 1979. Pattern and process in a forested ecosystem. Springer-Verlag, New York. Briske, D. D., and A. M. Wilson. 1977. Temperature effects on adventitious root development in blue grama seedlings. Journal of Range Management 30:276–280. Briske, D. D., and A. M. Wilson. 1978. Moisture and temperature requirements for adventitious root development in blue grama seedlings. Journal of Range Management 31:174–178. Burke, I. C., W. K. Lauenroth, and D. P. Coffin. 1995. Soil organic matter recovery in semiarid grasslands: Implications for the Conservation Reserve Program. Ecological Applications 5:793–801. Burke, I. C., W. K. Lauenroth, R. Riggle, P. Brannen, B. Madigan, and S. Beard. 1999. Spatial variability of soil properties in the shortgrass steppe: The relative importance of topography, grazing, microsite, and plant species in controlling spatial patterns. Ecosystems 2:422–438. Burke, I. C., C. M. Yonker, W. J. Parton, C. V. Cole, K. Flach, and D. S. Schimel. 1989. Texture, climate, and cultivation effects on soil organic matter content in U.S. grassland soils. Soil Science Society of America Journal 53:800–805. Carlson, J. M., and T. O. Crist. 1999. Plant responses to pocket-gopher disturbances across pastures and topography. Journal of Range Management 52:637–645. Carlson, D. C., and E. M. White. 1987. Effects of prairie dogs on mound soils. Soil Science Society of America Journal 51:389–393. Clark, W. H., and P. L. Comanor. 1975. Removal of annual plants from the desert ecosystem by Western harvester ants, Pogonomyrex occidentalis. Environmental Entomology 4:52–56. Clements, F. E. 1916. Plant succession: An analysis of the development of vegetation. Publication no. 242. Carnegie Institute of Washington, Washington, D.C. Clements, F. E. 1928. Plant succession and indicators: A definitive edition of plant succession and plant indicators. H. W. Wilson, New York. Coffin, D. P., and W. K. Lauenroth. 1988. The effects of disturbance size and frequency on a shortgrass plant community. Ecology 69:1609–1617. Coffin, D. P., and W. K. Lauenroth. 1989a. Disturbances and gap dynamics in a semiarid grassland: A landscape-level approach. Landscape Ecology 3(1):19–27. Coffin, D. P., and W. K. Lauenroth. 1989b. Small scale disturbances and successional dynamics in a shortgrass community: Interactions of disturbance characteristics. Phytologia 67(3):258–286. Coffin, D. P., and W. K. Lauenroth. 1989c. The spatial and temporal variability in the seed bank of a semiarid grassland. American Journal of Botany 76(1):53–58. Coffin, D. P., and W. K. Lauenroth. 1990a. A gap dynamics simulation model of succession in the shortgrass steppe. Ecological Modelling 49:229–266. Coffin, D. P., and W. K. Lauenroth. 1990b. Vegetation associated with nest sites of Western harvester ants (Pogonomyrex occidentalis) in a semiarid grassland. American Midland Naturalist 123:226–235.

The Role of Disturbances in Community and Ecosystem Dynamics 115 Coffin, D. P., and W. K. Lauenroth. 1992. Spatial variability in seed production of the perennial bunchgrass Bouteloua gracilis (H.B.K.) Lag. ex Griffiths. American Journal of Botany 79:347–353. Coffin, D. P., and W. K. Lauenroth. 1994. Successional dynamics of a semiarid grassland: Effects of soil texture and disturbance size. Vegetatio 110:67–82. Coffin, D. P., W. K. Lauenroth, and I. C. Burke. 1993. Spatial dynamics in recovery of shortgrass steppe ecosystems. Lectures on Mathematics in the Life Sciences 23:75–108. Coffin, D. P., W. K. Lauenroth, and I. C. Burke. 1996. Recovery of vegetation in a semiarid grassland 53 years after disturbance. Ecological Applications 6:538–555. Coffin, D. P., W. A. Laycock, and W. K. Lauenroth. 1998. Disturbance intensity and above- and belowground herbivory effects on long-term (14y) recovery of a semiarid grassland. Plant Ecology 139:221–233. Coppock, D. L., J. K. Detling, J. E. Ellis, and M. I. Dyer. 1983. Plant–herbivore interactions in a North American mixed-grass prairie. I. Effects of black-tailed prairie dogs on intraseasonal aboveground plant biomass and nutrient dynamics and plant species diversity. Oecologia 56:1–9. Costello, D. F. 1944. Natural revegetation of abandoned plowed land in the mixed prairie association of northeastern Colorado. Ecology 25:312–326. Crist, T. O., and J. A. Wiens. 1994. Scale effects of vegetation on forager movement and seed harvesting by ants. Oikos 69:37–46. Crist, T. O., and J. A. Wiens. 1996. The distribution of ant colonies in a semiarid landscape: Implications for community and ecosystem processes. Oikos 76:301–311. Dahlsted, K. J., S. Sather-Blair, B. K. Worcester, and R. Klukas. 1981. Application of remote sensing to prairie dog management. Journal of Range Management 34:218–223. Detling, J. K., and E. L. Painter. 1983. Defoliation responses of western wheatgrass populations with diverse histories of prairie dog grazing. Oecologia 57:65–71. Fair, J. L., D. P. C. Peters, and W. K. Lauenroth. 2001. Response of Bouteloua gracilis (Gramineae) plants and tillers to small disturbances. American Midland Naturalist 145:147–158. Foster, M. A., and J. Stubbendieck. 1980. Effects of the plains pocket gopher (Geomys bursarius) on rangeland. Journal of Range Management 33:74–78. Fraleigh, H. D., Jr. 1999. Seed dispersal of two important perennial grasses in the shortgrass steppe. Masters thesis, Colorado State University, Fort Collins, Colo. Fresquez, P. R., R. Aguilar, R. E. Francis, and E. F. Aldon. 1991. Heavy metal uptake by blue grama growing in a degraded semiarid soil amended with sewage sludge. Journal of Water, Air, and Soil Pollution 57–58:903–912. Fresquez, P. R., R. E. Francis, and G. L. Dennis. 1990a. Effects of sewage sludge on soil and plant quality in a degraded semiarid grassland. Journal of Environmental Quality 19:324–329. Fresquez, P. R., R. E. Francis, and G. L. Dennis. 1990b. Soil and vegetation responses to sewage sludge on a degraded semiarid broom snakeweed/blue grama plant community. Journal of Range Management 43:325–331. Gleason, H. A. 1926. The individualistic concept of the plant association. Bulletin of the Torrey Botanical Club 53:7–26. Grant, W. E., N. R. French, and L. J. Folse, Jr. 1980. Effects of pocket gopher mounds on plant production in shortgrass prairie ecosystems. Southwestern Naturalist 25:215–224. Hansen, R. M., and I. K. Gold. 1977. Blacktail prairie dogs, desert cottontails and cattle trophic relations on shortgrass range. Journal of Range Management 30:210–214. Hartley, L. M. 2006. Plague and the black-tailed prairie dog: An introduced disease mediates the effects of an herbivore on ecosystem structure and function. PhD diss, Colorado State University, Fort Collins, Colo.

116

Ecology of the Shortgrass Steppe

Hook, P. B., I. C. Burke, and W. K. Lauenroth. 1991. Heterogeneity of soil and plant N and C associated with individual plants and openings in North American shortgrass steppe. Plant and Soil 138:247–256. Hook, P. B., and W. K. Lauenroth. 1994. Root system response of a perennial bunchgrass to neighborhood-scale soil water heterogeneity. Journal of Ecology 8:738–745. Hook, P. B., W. K. Lauenroth, and I. C. Burke. 1994. Spatial patterns of roots in a semiarid grassland: Abundance of canopy openings and regeneration gaps. Journal of Ecology 82:485–494. Horn, B. E., and E. F. Redente. 1998. Soil nitrogen and plant cover of an old-field on the shortgrass steppe in southeastern Colorado. Arid Soil Research and Rehabilitation 12:193–206. Hyder, D. N., and R. E. Bement. 1972. Controlling red threeawn on abandoned cropland with ammonium nitrate. Journal of Range Management 25:443–446. Hyder, D. N., and A. C. Everson. 1968. Chemical fallow of abandoned croplands on the shortgrass prairie. Weed Science 16:531–533. Hyder, D. N., A. C. Everson, and R. E. Bement. 1971. Seedling morphology and seeding failures with blue grama. Journal of Range Management 24:287–292. Ihori, T., I. C. Burke, and P. B. Hook. 1995a. Nitrogen fertilization in native cultivated and abandoned fields in shortgrass steppe. Plant and Soil 171:203–208. Ihori, T., I. C. Burke, W. K. Lauenroth, and D. P. Coffin. 1995b. Effects of cultivation and abandonment on soil organic matter in northeastern Colorado. Soil Science Society of America Journal 59:1112–1119. Jaramillo, V. J., and J. K. Detling. 1988. Grazing history, defoliation, and competition: Effects on shortgrass production and nitrogen accumulation. Ecology 69:1599–1608. Judd, I. B. 1974. Plant succession on old fields in the Dust Bowl. Southwestern Naturalist 19:227–239. Judd, I. B., and M. L. Jackson. 1939. Natural succession of vegetation on abandoned farmland in the Rosebud soil area of western Nebraska. Journal American Society of Agronomy 39:541–547. Keeler, K. H. 1993. Fifteen years of colony dynamics in Pogonomyrex occidentalis, the Western harvester ants, in western Nebraska. Southwestern Naturalist 38:286–289. Kelly, R. H., and I. C. Burke. 1997. Heterogeneity of soil organic matter following death of individual plants in shortgrass steppe. Ecology 78:1256–1261. Kelly, R. H., I. C. Burke, and W. K. Lauenroth. 1996. Soil organic matter and nutrient availability responses to reduced plant inputs in shortgrass steppe. Ecology 77:2516–2527. Klatt, L. E., and D. Hein. 1978. Vegetative differences among active and abandoned towns of black-tailed prairie dogs (Cynomys ludovicianus). Journal of Range Management 31:315–317. Klipple, G. E., and D. F. Costello. 1960. Vegetation and cattle responses to different intensities of grazing on shortgrass ranges of the Central Great Plains. USDA–ARS technical bulletin no. 1216. U.S. Department of Agriculture, Washington, D.C. Knowles, C. J. 1986. Some relationships of black-tailed prairie dogs to livestock grazing. Great Basin Naturalist 46:198–203. Lathrop, E. 1982. Recovery of perennial vegetation in military maneuver areas, pp. 269–276. In: R. H. Webb and H. G. Wilshire (eds.), Environmental effects of off-road vehicles: Impacts and management in arid regions. Springer-Verlag, New York. Lauenroth, W. K., and D. P. Coffin. 1992. Belowground processes and the recovery of semiarid grasslands from disturbance, pp. 131–150. In: M. K. Wali (ed.), Ecosystem rehabilitation: Preamble to sustainable development. Vol. 2: Ecosystem analysis and synthesis. SPB Academic Publishing, The Hague, Netherlands.

The Role of Disturbances in Community and Ecosystem Dynamics 117 Lauenroth, W. K., J. L. Dodd, and P. L. Sims. 1978. The effects of water- and nitrogeninduced stresses on plant community structure in a semiarid grassland. Oecologia 36:211–222. Lavigne, R. J. 1969. Bionomics and nest structure of Pogonomyrex occidentalis (Hymenoptera: Formicidae). Annals Entomological Society of America 62:1166–1175. Laycock, W. A. 1989. Secondary succession and range condition criteria: Introduction to the problem, pp. 1–15. In: W. K. Lauenroth and W. A. Laycock (eds.), Secondary succession and the evaluation of rangeland condition. Westview Special Studies in Agriculture Science and Policy. Westview Press, Boulder, Colo. Laycock, W. A. 1991. Stable steady states and thresholds of range condition on North American rangelands: A viewpoint. Journal of Range Management 44:427–433. Martinez-Turanzas, G., D. P. Coffin, and I. C. Burke. 1997. Development of microtopographic relief in a semiarid grassland: Effects of disturbance size and soil texture. Plant and Soil 191:163–171. Milchunas, D. G., and W. K. Lauenroth. 1995. Inertia in plant community structure: State changes after cessation of nutrient-enrichment stress. Ecological Applications 5:452–458. Milchunas, D. G., K. A. Schultz, and R. B. Shaw. 1999. Plant community responses to disturbance by mechanized military maneuvers. Journal of Environmental Quality 28:1533–1547. Milchunas, D. G, W. K. Lauenroth, P. L. Chapman, and M. K. Kazempour. 1990. Community attributes along a perturbation gradient in a shortgrass steppe. Journal of Vegetation Science 1:375–384. Noy-Meir, I. 1973. Desert ecosystems: Environment and producers. Annual Review of Ecology and Systematics 4:25–51. O’Meilia, M. E., F. L. Knopf, and J. C. Lewis. 1982. Some consequences of competition between prairie dogs and beef cattle. Journal of Range Management 35:580–585. Pickett, S. T. A., and P. S. White. 1985. The ecology of natural disturbance and patch dynamics. Academic Press, New York. Reichhardt. K. L. 1982. Succession of abandoned fields on the shortgrass prairie, northeastern Colorado. Southwestern Naturalist 27:299–304. Riegel, A. 1941. Life history habits of blue grama. Kansas Academy of Sciences Transactions 44:76–83. Rogers, L. E. 1972. The ecological effects of the Western harvester ant (Pogonomyrex occidentalis) in the shortgrass prairie ecosystem. PhD diss., University of Wyoming, Laramie, Wyo. Rogers, L. E. 1974. Foraging activity of the Western harvester ant in the shortgrass plains ecosystem. Environmental Entomology 3:420–424. Rogers, L. E., and R. J. Lavigne. 1974. Environmental effects of Western harvester ants on the shortgrass plains ecosystem. Environmental Entomology 3:994–997. Sala, O. E., and W. K. Lauenroth. 1982. Small rainfall events: An ecological role in semiarid regions. Oecologia 53:301–304. Sala, O. E., W. K. Lauenroth, and W. J. Parton. 1981. Plant recovery following prolonged drought in a shortgrass steppe. Agricultural Meteorology 27:49–58. Samuel, M. J. 1985. Growth parameter differences between populations of blue grama. Journal of Range Management 38:339–342. Savage, D. A., and H. E. Runyon. 1937. Natural revegetation of abandoned farmland in the central and southern Great Plains, pp. 178–182. In: Grassland ecology. Section 1, Fourth International Grassland Congress, Aberystwyth, UK. Schwartz, C. C. 1977. Pronghorn grazing strategies on the shortgrass prairie, Colorado. PhD diss., Colorado State University, Fort Collins, Colo.

118

Ecology of the Shortgrass Steppe

Senft, R. L. 1983. The redistribution of nitrogen by cattle. PhD diss., Colorado State University, Fort Collins, Colo. Shantz, H. L. 1917. Plant succession on abandoned roads in eastern Colorado. Journal of Ecology 5:19–42. Shaw, R. B., and V. E. Diersing. 1990. Tracked vehicle impacts on vegetation at the Pinon Canyon Maneuver site, Colorado. Journal of Environmental Quality 19:234–243. Stanton, N. L., D. Morrison, and W. A. Laycock. 1984. The effect of phytophagous nematode grazing on blue grama die-off. Journal of Range Management 37:447–450. Ubbelohde, C., M. Benson, and D. A. Smith. 1988. A Colorado history. Pruett, Boulder, Colo. Ueckert, D. N. 1979. Impact of a white grub (Phyllophaga crinita) on a shortgrass community and evaluation of selected rehabilitation practices. Journal of Range Management 32:445–448. Vaughan, T. A. 1967. Food habits of the northern pocket gopher on shortgrass prairie. American Midland Naturalist 77:176–189. Vinton, M. A., and I. C. Burke. 1995. Interactions between individual plant species and soil nutrient status in shortgrass steppe. Ecology 76:1116–1133. Watt, A. S. 1947. Pattern and process in the plant community. Journal of Ecology 35:1–22. Weaver, J. E., and I. M. Mueller. 1942. Role of seedlings in recovery of midwestern ranges from drought. Ecology 23:275–294. Weiner, L. F., and J. L. Capinera. 1980. Preliminary study of the biology of the white grub Phyllophaga fi mbripes (LeConte) (Coleoptera: Scarabaeidae). Journal of the Kansas Entomological Society 53:701–710. Whicker, A. D., and J. K. Detling. 1988. Ecological consequences of prairie dog disturbances. BioScience 38:778–785. White, E. M., and D. C. Carlson. 1984. Estimating soil mixing by rodents. Proceedings South Dakota Academy Science 63:34–37.

7 Simulation of Disturbances and Recovery in Shortgrass Steppe Plant Communities Debra P. C. Peters William K. Lauenroth

S

imulation modeling is a complementary tool to field observation and experimentation in understanding ecological systems (Lauenroth et al., 1998). The overall objective of our plant community modeling is to allow us to evaluate the importance of gap dynamics concepts of succession for understanding shortgrass plant community recovery after disturbances. A gap dynamics approach focuses on individual plants, and the interactions between disturbance characteristics and plant life history traits in explaining successional patterns (Watt, 1947). Simulation models have been used extensively to evaluate the importance of gap dynamics processes to short- and long-term vegetation dynamics in temperate and tropical forests (e.g., Botkin et al., 1972; Shugart, 1984). We developed a gap dynamics model for shortgrass steppe plant communities (STEPPE [Coffin and Lauenroth, 1990]) based upon the conceptual and modeling framework provided by forest models, modifying it to represent Great Plains grasslands (Coffin and Lauenroth, 1996; Coffin and Urban, 1993). We used STEPPE in several capacities: (1) to synthesize and integrate existing knowledge to improve our understanding of recovery processes after disturbance, (2) to identify key processes limiting recovery, and (3) to predict long-term recovery dynamics for different climate and disturbance characteristics—in particular, soil texture and disturbance size. Our approach to modeling shortgrass community dynamics was to incorporate only the most important processes needed to address specific research questions. We added processes through time either because the model did not sufficiently represent ecosystem dynamics or because we posed more complicated research questions.

119

120

Ecology of the Shortgrass Steppe

STEPPE Model Description STEPPE simulates the recruitment, growth, and mortality of individual plants on a small plot through time at an annual time step (Fig. 7.1) (Coffin and Lauenroth, 1990). Recruitment and mortality both have stochastic elements. Growth is deterministic and is based upon competition for resources among plants. A key difference between STEPPE and the forest models from which it was derived is that belowground resources are the most frequently limiting resources in semiarid grasslands compared with aboveground resources (light) in forests (Lauenroth and Coffin, 1992). Because of the importance of Bouteloua gracilis to shortgrass communities, we hypothesized that the death of a full-size B. gracilis plant results in a gap in belowground resource space that initiates the successional processes of gap dynamics. Thus, the simulated plot size (0.12 m2) is based upon the belowground resource space associated with the roots of a full-size B. gracilis plant (Coffin and Lauenroth, 1991). The model simulates 15 groups of species chosen to represent the life history traits of the more than 300 species found in shortgrass communities at the SGS LTER site. These groups are further aggregated into five resource groups based on spatial distributions of root biomass. Size and age of each plant on a plot are simulated at an annual time step. Driving variables include precipitation, temperature, and soil texture. Model parameters and driving variables are obtained primarily from studies conducted at the SGS LTER site. The model can simulate

STEPPE (individual plant model)

Growth 1. soil water 2. temperature 3. nitrogen 4. plant interactions 5. plant size

Recruitment 1. soil water 2. temperature 3. site conditions 4. propagule density

Mortality 1. disturbances 2. life span 3. suppressed growth

35 cm Figure 7.1 The STEPPE, individual plant-based gap dynamics model for shortgrass plant communities.

Simulation of Disturbances and Recovery in Plant Communities 121

either a single plot or a grid of plots that interact through seed dispersal. Because the model is stochastic, a large number of replicate plots or grids (25–50) are typically simulated, and the results are averaged for each time step.

Recruitment Each year there is a probability of establishment of either a seedling or a vegetative propagule for each species. In single-plot simulations, we assume that seeds of all species are present on a plot and resources are available for establishment every year. In grid simulations, seed availability is dependent upon proximity to plots on which seeds are produced. The probability that a seedling from a particular species will become established is based either on the occurrence of suitable microenvironmental conditions or on the relative abundance of seeds on a plot. Establishment of the dominant species, B. gracilis, is based on the probability (0.125⋅y–1) that microenvironmental conditions for establishment would occur each year (Lauenroth et al., 1994). For all other species, the probability of establishment is based on the relative abundance of seeds produced, which is estimated from seed production data collected at the SGS LTER site. We assume that one to five species establish each year, with one to three seedlings added for each species. The exact number of species and seedlings is determined by drawing random numbers. All seedlings are added at the estimated size of a 1-year-old plant. Vegetative propagation occurs in clonal species or those with deep tap roots. We assume that plants have a 75% to 90% chance of regrowth after death of aboveground parts, depending on the source of mortality. If vegetative propagation occurs, then one to three plants of that species are added to the plot.

Mortality The probability of mortality for each current individual is determined by the disturbance rate, the longevity of the species, and a risk of death associated with slow-growing plants. Effects of cattle fecal pats, western harvester ant mounds, and burrows from small animals are incorporated into STEPPE using their frequencies of occurrence, as reported in Coffin and Lauenroth (1988). Similar to forest gap models, we assume that each species or group has an age-independent intrinsic likelihood of mortality, such that only 1% of a cohort growing under optimal conditions will reach maximum age. We also assume that slow-growing plants have a greater risk of death from disease, insects, and severe environmental conditions than do plants having average growth rates. This probability of mortality (0.368⋅y–1) results in slow-growing plants having a 1% chance of surviving 10 years. Clonal plants, such as B. gracilis and Opuntia polyacantha, are excluded from this source of mortality. Mortality of B. gracilis clumps occurs as a result of insufficient resources or of disturbances. Opuntia polyacantha plants experience these same sources of mortality with high growing season precipitation being an additional source (Dodd and Lauenroth, 1975).

122

Ecology of the Shortgrass Steppe

Growth Growth of plants occurs annually as a function of optimum growth rates, effects of precipitation and temperature, and interactions with other plants for belowground resources. An optimum growth rate estimated from the literature for each species is used to calculate the amount of resources required for each plant in a resource group to grow optimally. Soil water and interactions with other plants determine the actual amount of resources available to each group. STEPPE associates a particular proportion of belowground resource space with individual species or groups of species based on root distribution with depth, distribution of resources with depth, and the temporal variation in both distributions. Root distributions by depth are estimated from the literature and field data. Because soil water is the most frequent control on plant growth and community structure in semiarid grasslands (Lauenroth et al., 1978; Noy-Meir, 1973), the distribution of resources is based on soil water availability with depth in the soil profile. We have two different methods of determining soil water availability, depending upon whether STEPPE is run alone or run in conjunction with SOILWAT, a daily time step soil water simulation model (Parton, 1978; Sala et al., 1992). In the standalone version, temporal variation in water availability to each species depends on both the amount of annual precipitation and the size of plants of a particular species relative to the size of other plants. In the SOILWAT–STEPPE simulations, we integrate daily soil water values from SOILWAT to determine the amount of water available to each species on an annual basis. This is elaborated in the section on STEPPE–SOILWAT simulations. Regardless of the source of information about soil water availability, the actual growth rate for a plant is based on the relationship between resources required for optimum growth and resources available to the plant.

Single-Plot Simulations We conducted a set of single-plot simulations to test whether processes that result in plot–plot interactions are important for determining the structure of shortgrass steppe plant communities. Single-plot simulations assume that each plot is independent of all plots that surround it. This means that propagules of each species are always available from either the seedbank or from propagule production. Our assumption is appropriate for an isolated single gap disturbance in which the plot is surrounded by undisturbed vegetation. STEPPE can be used in this mode to evaluate successional dynamics and the time required for B. gracilis to dominate total plant cover after disturbance (Coffin and Lauenroth, 1990). Disturbance rates for the SGS LTER sites are based on frequencies of occurrence of cattle fecal pats, nest sites of Western harvester ants, and burrows from small animals (Coffin and Lauenroth, 1988; Peters et al., chapter 6, this volume ). In our initial approach to conducting these simulations, we further assumed that the occurrence of a disturbance killed all plants on the plot, and that all species had an equal probability of establishing from seed. In single-plot mode, we found that although basal cover on the simulated plot is eventually dominated by B. gracilis, during the initial 20 years after a

Aboveground biomass (g m -2)

Simulation of Disturbances and Recovery in Plant Communities 123

(A)

100 80

B. gracilis Perennial forbs/subshrubs Perennial graminoids

60 40 20 0

Aboveground biomass (g m -2)

0

40

120

80

160

200

240

(B)

100 80

Succulents Annuals

60 40 20 0 0

40

80

120

160

Time (years)

200

240

Figure 7.2 STEPPE results for the average of 50 plots for 250 years assuming seeds always available for all species: aboveground biomass of B. gracilis, perennial graminoids, and perennial forbs and shrubs (A); and aboveground biomass of annuals and succulents (B). (Coffin and Lauenroth [1990a].)

disturbance other perennial graminoids are the dominants (Fig. 7.2). Although the relative proportions of aboveground biomass and average biomass predicted for each species group by these simulations are comparable with the composition of shortgrass communities in northeastern Colorado (Coffin and Lauenroth, 1989b), the time required for B. gracilis to recover is much faster in the simulations than has been reported from field studies. Field and laboratory experiments indicated that a restrictive set of temperature and soil water conditions are required for B. gracilis seed germination and establishment (Briske and Wilson, 1977, 1978), suggesting that our assumption about equal probability of seedling establishment is incorrect. Using historical weather and Monte Carlo simulations with SOILWAT, we estimated the probability of appropriate conditions for seedling establishment of B. gracilis (Lauenroth et al., 1994). Using this probability of establishment, simulation prediction of average recovery time for B. gracilis after a disturbance increased from 20 to 65 years, which is approximately what has been found in field studies (Fig. 7.3). The problem with these results is that the model predicts the exact same recovery time and dynamics regardless of the size of the disturbance. Observations from abandoned agricultural fields make it clear that this prediction of recovery time is too fast, and suggests that other processes have important influences on recovery at scales larger than a single gap.

124

Ecology of the Shortgrass Steppe

Aboveground biomass (g m -2)

100 80 B. gracilis Perennial graminoids Perennial forbs and subshrubs

60 40 20 0 0

50

100

150

200

250

Time (years) Figure 7.3 Simulated aboveground biomass for the average of 50 plots for 250 years under the condition that B. gracilis seeds have a probability less than 1.0 of being present on the plot. (Coffin and Lauenroth [1990a].)

Spatially Explicit Simulations Seed Dispersal The discrepancy between our simulated recovery time of B. gracilis on an individual plot and the predicted time based upon observations of old fields led us to hypothesize that spatially explicit processes are important as disturbance size increases beyond a single gap. We incorporated spatial structure into STEPPE by considering that each disturbed area consists of a grid of plots in which processes on one plot could affect processes on all neighboring plots (Coffin and Lauenroth, 1989a). The key spatially explicit process we added to the model was seed dispersal. In addition to constraints on germination and establishment, in these simulations the availability of B. gracilis seeds to each plot is based on probabilities associated with the production and dispersal of seeds. Seed production is assumed to be a function of B. gracilis biomass on each plot and the amount of annual precipitation received the previous year (Coffin and Lauenroth, 1992). Because storage of B. gracilis seeds in the soil is low and variable through time and space (Coffin and Lauenroth, 1989c), we assumed that seed availability is primarily a function of seeds produced during the previous year. The probability of seeds dispersing to each plot is assumed to be a function of the distance from the nearest source of seeds, the release height of seeds from the inflorescence, average wind speed, and the aerodynamic properties of B. gracilis seeds. Simulations of disturbances using the spatially explicit version for sizes ranging from 2 to 16 m2 show that recovery time of B. gracilis increases as disturbance

Biomass (g m-2)

Biomass (g m-2)

Simulation of Disturbances and Recovery in Plant Communities 125 120 100 80 60 40 20 0 -20

(A)

Independent Dependent C. bands

100

(B)

80 60 40 20

Biomass (g m-2)

0 100

(C)

80 60 40 20 0 0

40

80

120

160

200

Time (years)

240

280

Figure 7.4 Simulated aboveground biomass of B. gracilis (average and 95% confidence intervals) for 300 years for two types of landscapes and three disturbance sizes: 2 m2 (A), 18 m2 (B), and 49 m2 (C). (From Coffin and Lauenroth [1989a]. Reprinted with kind permission of Springer Science and Business Media.)

size increases, and that estimates of recovery time are much greater from the spatially explicit than from the spatially independent version (Fig. 7.4 [Coffin and Lauenroth, 1989a]). Recovery in these spatially explicit simulations was defined as the time at which the 95% confidence interval of the simulated biomass includes the steady-state value predicted by the independent simulations (Fig. 7.4). The inclusion of spatial processes associated with seed dispersal increased the recovery time of B. gracilis because plots that are not within the maximum dispersal distance after the disturbance cannot receive B. gracilis seeds until nearby plots have been colonized. The time required for a particular plot to be colonized by B. gracilis is proportional to its distance from the closest plot that has sufficient B. gracilis to produce seeds. The most relevant unit for measuring this is D, the maximum dispersal distance. These spatially explicit recovery times are in line with results from studies of abandoned agricultural fields. Soil Texture Because disturbances occur over a range of soil textures, and soil properties have important effects on plant processes, we hypothesized that soil texture would influence the recovery of B. gracilis for different disturbance sizes (Coffin and

126

Ecology of the Shortgrass Steppe

Lauenroth, 1994). We used results from Monte Carlo simulations based on historical weather data and SOILWAT to determine the average annual probability of seedling establishment for five soil texture classes (Lauenroth et al., 1994). We found that the probability of establishment of B. gracilis increases as silt content of the soil increases. We simulated effects of soil texture in the same manner as effects of resource availability on rate of growth using a statistical relationship between annual precipitation, aboveground net primary production, and water-holding capacity of the soil (Sala et al., 1988). Our model results show that soil texture is more important than disturbance size to simulated recovery of B. gracilis, and constraints on recruitment are more important than constraints on growth (Coffin and Lauenroth, 1994). Fastest recovery occurs on soils with the highest silt content (silt loam); slowest recovery occurs on soils with low silt content and either high (clay) or low (loamy sand) water-holding capacity (Fig. 7.5). Biomass and recovery rate of B. gracilis decrease as disturbance size (2–16 m 2) increases and as distance from the disturbed plot to the edge of undisturbed vegetation increases.

80

(A) 2 m 2

60

Aboveground biomass (g m -2)

40 20 0 80

(B) 18 m 2

60 40 20 0 (C) 49 m 2

80

Figure 7.5 Simulated average aboveground biomass of B. gracilis for 500 years for three disturbance sizes and five soil texture classes. (From Coffin and Lauenroth [1994]. Reprinted with kind permission of Springer Science and Business Media.)

sandy clay loam loamy sand silt loam loam clay

60 40 20 0 0

100

200

300

Time (years)

400

500

Simulation of Disturbances and Recovery in Plant Communities 127

STEPPE–SOILWAT Simulations We used a spatially explicit modeling approach to link STEPPE with a model of soil water dynamics to simulate vegetation dynamics and recovery rates of B. gracilis on abandoned agricultural fields (Coffin et al., 1993). We evaluated the effects of seed dispersal, weather, soil texture, and nitrogen availability on recovery. Availability of soil water or nitrogen to each plant was simulated as a function of its root distribution with depth, the distribution of resources by depth, and temporal variation in the distributions. We represented plant growth and soil water interactions dynamically using a one-to-one correspondence between a plot simulated by STEPPE and one simulated by SOILWAT. Nitrogen availability was calculated within STEPPE based on a relationship between soil texture and time after abandonment, which was obtained by running simulations with the CENTURY soil process model (Parton et al., 1987; Parton et al., chapter 15, this volume). Plots were arrayed into a grid that allowed them to be interconnected to represent old fields. We simulated two sites in northeastern Colorado with different long-term precipitation and temperature. Within each site, we simulated fields with different soil textures. Our results show that simulated recovery patterns vary both between and within fields (Coffin et al., 1993). Variability in patterns between fields is the result of differences in soil texture, with the fastest recovery occurring on fields with silt loam soils (Fig. 7.6). Precipitation is less important than soil texture, even though the fastest recovery occurred on fields with the highest precipitation. Distance from the source of seeds and soil texture both have important effects on spatial patterns of recovery within each field. At any point in time, biomass of B. gracilis decreases as distance from the edge of a field increases. Biomass also decreases with decreasing silt content. Although the general pattern of decrease in B. gracilis cover with distance from the edge of old fields is represented by the model, our simulated rate of recovery is slow compared with observed patterns (Coffin et al., 1996). Our model was developed assuming that recovery is dependent solely upon wind dispersal of seeds from the undisturbed edge of the field. Thus, these old-field simulations indicate that a mode of dispersal that can operate at distances much greater than wind may be important to the recovery process. Fraleigh (1999) evaluated dispersal distances for B. gracilis seeds over a range of wind speeds up to 19 m⋅s–1 and found that 98% of the seeds are found with 8 m of the release point with only a small fraction (0.3%) found beyond 11 m. The maximum dispersal distance by wind in the model is 8 m. Increasing the maximum dispersal distance to 11 m did not speed up recovery enough to match the field data. Biotic factors, such as the dissemination of seeds on the fur of cattle or through consumption and subsequent deposition in fecal pats, are other possible candidates for long-distance seed dispersal into abandoned fields. Fraleigh (1999) found that B. gracilis seeds can be transported on the hair of cattle for distances of at least 100 m and probably much further. Fraleigh (1999) also found substantial numbers of B gracilis seeds, that could be germinated, in cattle fecal pats. Thus, we have concluded that cattle may be the important link to explain the speed of recovery on abandoned

128

Ecology of the Shortgrass Steppe

100

(A) silt loam

80 60 40

Aboveground biomass (g m -2)

20 0 100

(B) silt clay

80 60 40 20 0 100

(C) sandy clay loam

80

B. gracilis biomass at CPER at four times: 50 100 150 200

60 Figure 7.6 Simulated aboveground biomass of B. gracilis at the CPER by distance from the edge of field at four times after abandonment (50, 100, 150, 200 years) for three fields with different soil textures. (From Coffin et al. [1993].)

40 20 0

0

10

20

30

40

Distance from edge (m)

agricultural fields in the shortgrass region. Incorporation of the effects of cattle as dispersal agents is the next logical step in simulating the recovery of B. gracilis on large disturbances.

Summary and Conclusions Simulation modeling provides a powerful approach to synthesizing information about ecological properties and processes to identify key controls on community dynamics and to predict long-term recovery patterns after disturbance.

Simulation of Disturbances and Recovery in Plant Communities 129

Our analyses in the shortgrass steppe show that seed dispersal is a key process limiting recovery of B. gracilis across a range of disturbance sizes. Our simulation results also indicate the importance of soil properties, especially silt content, to recovery rates of B. gracilis, and the greater importance of seedling establishment compared with seedling growth; however, these hypotheses have yet to be tested in the field. For large disturbances, the decrease in B. gracilis cover with increasing distance from the source of seeds at the field edge agrees with field and simulation results from small- and intermediate-size disturbances (Coffin and Lauenroth, 1989a, b; 1994). However, our simulated recovery times for large disturbances are longer than for small disturbances, and recovery is not a simple linear function of disturbance size. As disturbance size increases, surface area increases faster than the effective radius over which seeds are dispersed, with a corresponding exponential decrease in probability of seed dispersal into the disturbed area. High between-field variation observed by sampling 13 fields in northeastern Colorado (Coffin et al., 1996) was not captured by this modeling exercise, indicating the importance of other factors not currently included in the model. Simulation modeling will remain a powerful tool for understanding shortgrass steppe plant communities in the future as the global environment changes and temperatures continue to increase with increases in atmospheric CO2. In many cases, the ecological consequences of a changing climate are unknown, and the direction and amount of change in precipitation has a high degree of uncertainty for semiarid regions. Models can be used to explore future dynamics under different climate scenarios that include changes in both the amount and timing of precipitation as well as interactions with changes in temperature for a range of soil textures and disturbance sizes and types. These multifactorial, long-term simulations are easy to conduct using models, yet are very challenging and in some cases impossible to examine using experiments. Combining model results with focused short- and long-term experiments and observations will provide new insights into complex ecosystem dynamics.

Acknowledgments This research was supported by an NSF grant (BSR 9011659) to CSU as part of the SGS LTER program. We thank Brandon Bestelmeyer for assistance with the figures, and Tom Crist and one anonymous reviewer for helpful comments on the manuscript.

References Botkin, D. B., J. F. Janak, and J. R. Wallis. 1972. Some ecological consequences of a computer model of forest growth. Journal of Ecology 60:849–873. Briske, D. D., and A. M. Wilson. 1977. Temperature effects on adventitious root development in blue grama seedlings. Journal of Range Management 30:276–280. Briske, D. D., and A. M. Wilson. 1978. Moisture and temperature requirements for adventitious root development in blue grama seedlings. Journal of Range Management 31:174–178.

130

Ecology of the Shortgrass Steppe

Coffin, D. P., and W. K. Lauenroth. 1988. The effects of disturbance size and frequency on a shortgrass plant community. Ecology 69:1609–1617. Coffin, D. P., and W. K. Lauenroth. 1989a. Disturbances and gap dynamics in a semiarid grassland: A landscape-level approach. Landscape Ecology 3(1):19–27. Coffin, D. P., and W. K. Lauenroth. 1989b. Small scale disturbances and successional dynamics in a shortgrass community: Interactions of disturbance characteristics. Phytologia 67(3):258–286. Coffin, D. P., and W. K. Lauenroth. 1989c. The spatial and temporal variability in the seed bank of a semiarid grassland. American Journal of Botany 76(1):53–58. Coffin, D. P., and W. K. Lauenroth. 1990. A gap dynamics simulation model of succession in the shortgrass steppe. Ecological Modelling 49:229–266. Coffin, D. P., and W. K. Lauenroth. 1991. Effects of competition on spatial distribution of roots of blue grama. Journal of Range Management 44:67–70. Coffin, D. P., and W. K. Lauenroth. 1992. Spatial variability in seed production of the perennial bunchgrass Bouteloua gracilis (H.B.K.) Lag. ex Griffiths. American Journal of Botany 79:347–353. Coffin, D. P., and W. K. Lauenroth. 1994. Successional dynamics of a semiarid grassland: Effects of soil texture and disturbance size. Vegetatio 110:67–82. Coffin, D. P., and W. K. Lauenroth. 1996. Regional analysis of transient responses of grasslands to climate change. Climate Change 34:269–278. Coffin, D. P., W. K. Lauenroth, and I. C. Burke. 1996. Recovery of vegetation in a semiarid grassland 53 years after disturbance. Ecological Applications 6:538–555. Coffin, D. P., W. K. Lauenroth, and I. C. Burke. 1993. Spatial dynamics in recovery of shortgrass steppe ecosystems. Lectures on Mathematics in the Life Sciences 23:75–108. Coffin, D. P., and D. L. Urban. 1993. Implications of natural history traits to systemlevel dynamics: Comparisons of a grassland and a forest. Ecological Modelling 67:147–178. Dodd, J. L., and W. K. Lauenroth. 1975. Responses of Opuntia polyacantha to water and nitrogen perturbations in the shortgrass prairie, pp. 229–240. In: M. K. Wali (ed.), Prairie: A multiple view. University of North Dakota, Grand Forks, N. Dak. Fraleigh, H. D., Jr. 1999. Seed dispersal of two important perennial grasses in the shortgrass steppe. Masters thesis, Colorado State University, Fort Collins, Colo. Lauenroth, W. K., C. D. Canham, A. P. Kinzig, K. A. Poiani, W. M. Kemp, and S. W. Running. 1998. Simulation modeling in ecosystem science, pp. 404–415. In: M. L. Pace and P. M. Groffman (eds.), Successes, limitations and frontiers in ecosystem science. Springer-Verlag, New York. Lauenroth, W. K., and D. P. Coffin. 1992. Belowground processes and the recovery of semiarid grasslands from disturbance, pp. 131–150. In: M. K. Wali (ed.), Ecosystem rehabilitation. Vol 2. Ecosystem analysis and synthesis. SPB Academic Publishing, The Hague, the Netherlands. Lauenroth, W. K., J. L. Dodd, and P. L. Sims. 1978. The effects of water- and nitrogeninduced stresses on plant community structure in a semiarid grassland. Oecologia 36:211–222. Lauenroth, W. K., O. E. Sala, D. P. Coffin, and T. B. Kirchner. 1994. The importance of soil water in the recruitment of Bouteloua gracilis in the shortgrass steppe. Ecological Applications 4:741–749. Noy-Meir, I. 1973. Desert ecosystems: Environment and producers. Annual Review of Ecology and Systematics 4:25–51.

Simulation of Disturbances and Recovery in Plant Communities 131 Parton, W. J. 1978. Abiotic section of ELM, pp. 31–53. In: G. S. Innis (ed.), Grassland simulation model. Springer-Verlag, New York. Parton, W. J., D. S. Schimel, C. V. Cole, and D. S. Ojima. 1987. Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Science Society of America Journal 51:1173–1179. Sala, O. E., W. K. Lauenroth, and W. J. Parton. 1992. Long term soil water dynamics in the shortgrass steppe. Ecology 73:1175–1181. Sala, O. E., W. J. Parton, L. A. Joyce, and W. K. Lauenroth. 1988. Primary production of the central grassland region of the United States. Ecology 69:40–45. Shugart, H. H. 1984. A theory of forest dynamics. Springer-Verlag, New York. Watt, A. S. 1947. Pattern and process in the plant community. Journal of Ecology 35:1–22.

8 Ecology of Mammals of the Shortgrass Steppe Paul Stapp Beatrice Van Horne Mark D. Lindquist

A

t first glance, the shortgrass steppe seems to offer little in the way of habitat for mammals. The expansive rolling plains, with little topographic relief or vegetative cover, provide minimal protection from predators or the harsh weather typical of the region. The short stature of the dominant native grasses prevents the development of any significant litter layer, and although snowfall can often be significant, too little accumulates to form the subnivean habitats that support small mammal populations in forests and more productive grasslands in winter. As a consequence, ecologists have typically considered the vertebrate fauna of the shortgrass steppe to be depauperate compared with other Great Plains grasslands, a hardy collection of generalists living in sparse populations. Although this characterization may generally be accurate, it has led mammalian ecologists to overlook the fauna of the shortgrass steppe in favor of that of other grasslands. It is precisely these circumstances, however, that suggest that a long-term approach may be necessary to understand the dynamics of mammal populations here. Relatively few such studies have been completed to date, but we can use the comparative and experimental results that are available to begin to determine what factors might be important. Here we review research on mammals in the shortgrass steppe, with the goal of identifying the general patterns and processes that contribute to them. Our review is roughly divided into four parts. We begin by describing the mammal communities and their broad habitat associations in shortgrass steppe environments. We then review the history of mammal research in the region to synthesize what these studies (many unpublished) have taught us about the most important determinants of the distribution and abundance of native species. Studies of mammal 132

Ecology of Mammals of the Shortgrass Steppe 133

populations in the northern shortgrass steppe have spanned nearly 40 years, and we next describe some major patterns that have emerged from studies during this period. Much of this past research focused on the role of mammals in the structure and function of shortgrass steppe ecosystems, and we revisit this issue in some detail, with special emphasis on the important and sometimes controversial role of prairie dogs and other burrowing rodents. Finally, we end by considering how humans, and especially agriculture and its related activities, affect the diversity, abundance, and persistence of resident mammal populations. We make three caveats at the outset. First, most of the long-term research on shortgrass steppe mammals, as well as our own work, was conducted as part of Grassland Biome studies of the U.S. IBP and the NSF SGS LTER project. Our review thus emphasizes results from research at a single site in north–central Colorado, which currently encompasses only about 23% of the environmental variation in the shortgrass steppe region (Burke and Lauenroth, 1993), and thus may not be representative of all shortgrass steppe ecosystems. Second, because of their relative abundance and ease of capture and/or detection, many of the studies conducted during these programs focused on small mammals, principally rodents and rabbits. Although some information is available on species of economic importance (i.e., carnivores and ungulates), our research and this review reflects a bias toward small mammals. Last, studies of the shortgrass steppe have focused explicitly on areas with grassland cover. We advocate a broader view of the shortgrass steppe as a mosaic of habitat types, which includes shrublands, riparian zones, escarpments associated with permanent streams, and row–crop agricultural lands. Several mammals, including most bats, shrews, and species associated with eastern grasslands (e.g., eastern cottontails [Sylvilagus floridanus], eastern woodrats [Neotoma floridana], Virginia opossum [Didelphis virginiana]) are restricted largely to riparian areas and are not considered in detail here. However, most native species are found in multiple habitat types, all of which may be important for understanding population and community dynamics at landscape and regional scales.

The Shortgrass Steppe as a Habitat for Mammals As defined by Lauenroth and Milchunas (1991), the shortgrass steppe is located in the central Great Plains, bounded on the west by the Front Range of the Rocky Mountains and on all other sides by mixed-grass prairie or by desert. Geopolitically, the shortgrass steppe encompasses approximately the eastern third of both Colorado and New Mexico, the western Texas and Oklahoma panhandles, and southwestern Kansas, as well as small areas of southern Wyoming and Nebraska. The dominant plant communities have been described elsewhere (Lauenroth, chapter 5, this volume; Lauenroth and Milchunas, 1991), but a few points merit mention here because of their likely influence on mammal populations. First, with the exception of breaks and riparian strips near large waterways such as the South Platte, Arkansas, and Canadian rivers, most of the shortgrass steppe is characterized by rolling hills of shortgrass vegetation, dominated by two perennial grasses: blue grama (Bouteloua gracilis) and buffalograss (Buchloë

134

Ecology of the Shortgrass Steppe

dactyloides). Succulents (e.g., prickly pear [Opuntia polyacantha]), midgrasses, dwarf shrubs, and forbs are interspersed among grass plants, but most of the vegetation is short (0.2 m 2 in area were included. Soil are types based on Soil Conservation Survey map units. (P. Stapp, unpublished data.)

to be rare or absent in most sandy soils. Edaphic factors are also important for species that live aboveground, but these effects are mediated by the effects of soils on plant species and functional type diversity, and hence on habitat and food. Biotic Factors: Vegetation Structure As mentioned earlier, the availability of vegetative cover is probably the most significant single determinant of mammalian abundance, biomass, and diversity in the shortgrass steppe (Fig. 8.3). The presence of taller grasses, forbs, and shrubs alters small-mammal species composition and productivity by providing cover from predators and inclement weather, as well as nest materials for species such as harvest mice. At the SGS LTER site, rodent species diversity and density are approximately two times higher in areas with saltbush than without (Table 8.4), although total rodent biomass is roughly equivalent. Soils in the saltbush areas tend to be coarsely textured, and intermediate-height grasses such as Stipa comata, Agropyron smithii, and Oryzopsis hymenoides are also common. In addition to ground squirrels, four species of nocturnal rodents are captured regularly in saltbush areas, and prairie voles (M. ochrogaster) and hispid pocket mice (Chaetodipus hispidus) sometimes invade our trapping sites after wet years (P. Stapp, unpublished data). Stapp and Van Horne (1997) found that the population density of deer mice was correlated with the density and spatial pattern of shrubs in the shortgrass steppe. Mice tended to orient their movements toward shrubs in areas with few shrubs, but on sites with higher shrub canopy cover (>10%), mice showed no detectable preference for shrubs, possibly because they were able to achieve the protective benefits of shrub cover without actually moving beneath them. In contrast to saltbush sites, grasshopper mice and thirteen-lined ground squirrels are the only species captured consistently on upland shortgrass sites (Table 8.4). Grasshopper mice are more numerous in saltbush areas and their abundance increases in both shrub and grassland sites with the abundance of gopher mounds (Stapp, 1997b). Ground squirrels are found in both grassland and saltbush habitats,

144

Ecology of the Shortgrass Steppe

Figure 8.3 Areas with greater vegetative cover as such large shrubs (Atriplex canescens) and roadside verges are centers of activity and diversity. (Photo by Mark Vandever.)

but tend to be more abundant and have higher productivity in grasslands (Table 8.4). Black-tailed jackrabbits and desert cottontails favor areas with greater vegetative cover (Flinders and Hansen, 1975). More than 60% of black-tailed jackrabbits sighted during our roadside surveys were recorded in saltbush-dominated areas, even though saltbush represents only 27% of the vegetative cover along the survey route (P. Stapp, unpublished data). Desert cottontails tend to be most common near roadside ditches, in active and abandoned prairie dog colonies, and near livestock corrals.

Ecology of Mammals of the Shortgrass Steppe 145 Table 8.4 Rodent Population Densities (Measured in Number per Hectare) in Upland Prairie and Saltbush-Dominated Grasslands Species Northern grasshopper mouse (O/I) Deer mouse (O) Western harvest mouse (O/G) Ord’s kangaroo rat (G) Thirteen-lined ground squirrel (O) Nocturnal rodent species richness Mean rodent density, no./ha–1 Mean rodent biomass, g∙ha–1

Weight, g 33 21 12 69 124 — — —

Upland Prairie 0.92 0.17 0 0 3.69 1.84 4.81 492

Saltbush Grasslands 1.69 1.89 0.58 2.98 1.84 4.04 10.20 575

For nocturnal species, values are means based on captures from three 3.14-ha trapping webs in each habitat type sampled in May and September trapping sessions from 1994 to 2006, beginning in September 1994 (n = 25 trapping sessions). Each web consisted of 124 traps (12 lines of 10 traps at a 10-m spacing, with four traps at the center), which were set for four consecutive nights in each session. For ground squirrels, values are means of numbers of unique individuals captured in late June, prior to the emergence of young-of-year, from 1999 to 2006. Traps were set at 20-m intervals (62 traps/perweb) and checked for four consecutive mornings. For both nocturnal species and squirrels, density was calculated by dividing the number of unique individuals captured by web area (3.14 ha). Plains harvest mice, silky pocket mice (upland prairie), hispid pocket mice, and prairie voles (saltbush grasslands) are also occasionally captured but are rare. Letters in parentheses indicate trophic level: G, granivore; I, insectivore; O, omnivore.

Roadside vegetation is an important habitat for many rodents as well. These narrow habitats are only grazed or graded irregularly, and the overgrown vegetation, friable soils, and high seed densities support populations of rodents that are otherwise missing from shortgrass areas (Fig. 8.4 [Abramsky, 1978]). Roadside areas may also function as dispersal corridors and provide refugial habitats during periods of environmental stress. For example, deer mice were absent on all long-term trapping plots in late summer 1997, but a few individuals remained in roadside vegetation nearby. These individuals likely contributed to the recovery of deer mouse populations on plots the following year. In response to the high abundance of prey, carnivores such as swift foxes travel and hunt in roadside and other disturbed habitats (c.f., Cameron, 1984; Roell, 1999), and the availability of perches and roosts associated with roads and fence lines attract raptors such as owls (Zimmerman et al., 1996). For small mammals, life in roadside habitats therefore may reflect a tradeoff between higher cover and food availability and increased risk of encountering predators (Stapp and Lindquist, 2007). These comparative studies show the striking differences in small-mammal populations between areas that differ in vegetative cover. Experimental manipulations of nutrient and water stress demonstrated that changes in primary productivity of shortgrass vegetation also markedly affect small-mammal productivity and community structure. Between 1971 and 1974, Lauenroth et al. (1978) added nitrogen, water, and the two in combination to 1-ha plots, with two replicates of each treatment and two plots serving as controls, to determine effects of nutrient and water limitation on primary productivity and plant communities. At the same time, Grant et al. (1977) examined changes in small-mammal productivity and community structure in response to treatment-related changes in vegetation

146

Ecology of the Shortgrass Steppe

100

REME/MIOC PEMA DIOR ONLE

% of captures

80 60 40 20 0

0 -10

120

>200

Distance from fenceline (m) Figure 8.4 Frequency of captures of small mammals at different distances from weedy vegetation along fence lines adjacent to gravel roads in shrub-dominated areas of the shortgrass steppe. Small mammals were live-trapped on line transects placed at the fence line and 120 m into the pasture in 1997 and 1998. Abundance at more than 200 m was estimated concurrently on trapping webs as part of the SGS LTER population monitoring programs (P. Stapp, unpublished data). Values are percentage of captures per 100 trap nights on all areas. Species abbreviations: DIOR, Ord’s kangaroo rat; MIOC, prairie vole; ONLE, northern grasshopper mouse; PEMA, deer mouse; REME, western harvest mouse.

and arthropod productivity. Prairie voles and western harvest mice colonized the water and nitrogen-plus-water plots in response to increased plant biomass and invasion of these plots by weedy plants (Grant et al., 1977). Northern grasshopper mice, which prefer areas of bare soil and sparse vegetation (Egoscue, 1960; Kaufman and Fleharty, 1974), were mostly absent. Two dietary and habitat generalists, the deer mouse and thirteen-lined ground squirrel, were present on all treatments but responded differently; deer mice increased dramatically in number on the “wet” treatments (water, nitrogen plus water), whereas ground squirrels were significantly more abundant on the “dry” nitrogen and control plots. Grant et al. (1977) concluded that small-mammal community structure was closely tied to vegetative structure. Biotic Factors: Food Availability The results of the nutrient and water addition studies emphasized the importance of vegetative structure for small mammals, but the change in productivity might also have directly affected the abundance and quality of food. Vegetation and food availability are obviously linked for herbivores and granivores, but areas with greater structural complexity also support more arthropods and, hence, more prey for insectivores and omnivores. Grant et al. (1977) estimated that small

Ecology of Mammals of the Shortgrass Steppe 147

mammals consumed less than 4% of plant material and 34% of available arthropods, and concluded that neither herbivorous nor omnivorous rodent populations were limited by food availability. On a plot adjacent to the nitrogen and water experiment described earlier, Abramsky (1978) added alfalfa pellets and whole oats to examine the effects of food limitation directly. None of the resident species increased in number or biomass in response to supplemental food, which supports the conclusions of Grant et al. (1977). However, kangaroo rats colonized the food addition, apparently attracted by the novel food resource. Kangaroo rats are relatively uncommon in most areas of the shortgrass steppe, although a few individuals were actually captured earlier on the ecosystem stress plots (French and Grant, 1974). Like prairie voles and harvest mice, kangaroo rats likely dispersed onto plots from weedy vegetation along the adjacent roads, which underscores the importance of considering landscape context in interpreting results of these small-scale experiments. Abramsky’s (1978) experiment suggests that, of the resident small mammals, granivores such as kangaroo rats and pocket mice are one rodent group whose numbers may be closely tied to food availability. These species are uncommon in most shortgrass steppe areas except in sandy or disturbed soils, which support a greater diversity of forbs and midgrass species, and presumably a richer seedbank. Pocket gophers (T. talpoides) are another group whose distribution in the shortgrass steppe seems to be closely linked to its food source. Prickly pear cactus (O. polyacantha) makes up 50% of the annual diet of T. talpoides and may be particularly important in winter, when it was the dominant food eaten (79% of diet) and also an important source of water (Vaughan, 1967). A survey of the spatial distribution of gopher mounds and prickly pear cactus showed that gopher mounds are significantly associated with prickly pear—more than would be expected by chance (L. Dempsey and P. Stapp, unpublished data). Most studies of mammalian carnivores have focused largely on describing habitat relationships and diet composition, but have not evaluated possible factors that may limit their abundance in the shortgrass steppe. The density of mammalian predators (Table 8.5) is relatively low compared with other grasslands and shrub steppe ecosystems (e.g., Bekoff, 1982; Goodrich and Buskirk, 1998; Linzey, 1982; Messick and Hornocker, 1981), presumably as a result of the relatively low availability of vertebrate prey. Except for small mustelids such as skunks and

Table 8.5 Mean Population Density of Common Medium-Size and Large Mammals in the Shortgrass Steppe Species

Density, no./km 2

Coyote

0.13–0.17

Swift fox

0.07–0.23

American badger Pronghorn

0.21 0.25–0.67

References Flinders and Hansen (1975), Gese et al. (1989), Stapp (unpublished data) Covell (1992), Roell (1999), Rongstad et al. (1989), Stapp (unpublished data) Flinders and Hansen (1975) Pojar (1988), Pojar et al. (1995)

148

Ecology of the Shortgrass Steppe 4.0

3.0

14 12 10

2.5 8 2.0 6

1.5

Scat index

Rabbits seen km-1

3.5

all rabbits coyote scats

4

1.0

2

0.5 0.0 ‘94 ‘95 ‘96 ‘97 ‘98 ‘99 ‘00 ‘01 ‘02 ‘03 ‘04 ‘05 ‘06

0

Year of sampling Figure 8.5 Changes in relative abundance of rabbits and coyotes on the SGS LTER from spotlight and scat counts, respectively, along the same 32-km transect in north–central Colorado (P. Stapp, unpublished data). The scat index is the number of scats deposited per day per kilometer of transect × 100. Values are combined from surveys in spring (April/May) and autumn (October).

weasels, most of the resident carnivores are able to construct burrows (Fitzgerald et al., 1994) and thus are not obviously limited by habitat. Our monitoring studies show no close tracking of coyote populations with those of rabbits (Fig. 8.5), although these data are limited by the reliability of scat counts alone as an index of carnivore density (Schauster et al., 2002). Rabbit numbers typically peak in summer or autumn, whereas highest numbers of coyote scats are usually found in winter (Fig. 8.5), perhaps because of a slower rate of scat decomposition during winter months. Comprehensive studies of population dynamics of coyotes, swift foxes, and mustelids in the shortgrass steppe are badly needed. Biotic Factors: Species Interactions In a system with such low primary and secondary productivity, interactions among species arguably may be less significant in determining population densities than abiotic factors or the abundance of habitat or food. Still, in areas where habitat conditions permit coexistence, interactions can affect local population densities if individuals limit access to key resources such as food or cover. We know of no studies that have tested whether predation regulates or limits mammal populations in the shortgrass steppe. Predation by coyotes has been shown to be a major source of mortality for swift foxes (Kamler et al., 2003; Roell, 1999; Sovada et al., 1998). Peak rabbit numbers are often associated with periods of low coyote activity (Fig. 8.5), suggesting that rabbit numbers may be influenced by coyote abundance. In contrast, interspecific competition has been invoked regularly to

Ecology of Mammals of the Shortgrass Steppe 149

explain patterns of apparent niche segregation in grassland mammals, including the shortgrass steppe. Miller (1964), for example, interpreted the disjunct distributions of pocket gophers as the result of competition mediated by species-specific differences in preference for soil type. Moulton et al. (1983) disagreed with some of Miller’s (1964) conclusions and emphasized that changes in land use since the Dust Bowl may have led to competitive displacement of previously widespread species. Similarly, Burnett (1925 [cited in Fitzgerald et al., 1994]) suggested that conversion of native grasslands to agricultural and shrub-dominated vegetation favored black-tailed jackrabbits, which subsequently outcompeted native whitetailed jackrabbits (Flinders and Hansen, 1972). Throughout the 1970s and 1980s, small-mammal communities were widely used as experimental systems for studies of competition (Dueser et al., 1989), and one of the most widely cited pieces of evidence for the importance of competition arose from experiments conducted during the Grassland Biome studies. One unexpected outcome of the nitrogen and water addition “stress” experiment described earlier was the relatively small response of deer mice to changes in vegetative structure (Grant et al., 1977). Research from other systems (e.g., Grant, 1972; Redfield et al., 1977) has demonstrated that deer or white-footed mice often compete with voles for food and/or space, and Abramsky et al. (1979) speculated that deer mice and other “native” species avoided highly disturbed plots because of the competitive dominance of the invading species, particularly prairie voles. They subsequently removed both voles and harvest mice from one nitrogen-pluswater replicate, leaving the other replicate as a control. Deer mice increased in number after the removals, whereas grasshopper mice still avoided the nitrogenplus-water plots. Abramsky et al. (1979) concluded that voles excluded deer mice via exploitative or interference competition, whereas the avoidance of dense vegetation by grasshopper mice reflected habitat specialization that resulted from past competition. The studies by Abramsky et al. (1979) were conducted in an experimentally manipulated area of the shortgrass steppe (Abramsky, 1976, 1978; Abramsky and Tracy, 1979, 1980; Abramsky and Van Dyne, 1980), but interactions among rodent species may also be important in areas of native vegetation. Stapp (1997a) studied the role of competitive and predatory interactions between deer mice and grasshopper mice as determinants of local abundance of each species (Stapp, 1996). Unlike deer mice, which prefer shrub cover (Stapp and Van Horne, 1997), grasshopper mice show no affinity for shrubs and instead prefer open microhabitats (Stapp, 1997a). Grasshopper mice are known to prey opportunistically on other rodents, including deer mice (Bailey and Sperry, 1929; Flake, 1973; Rebar and Conley, 1983), and Stapp (1997a) speculated that one explanation for the preference of deer mice for shrubs, and their rarity in most grassland areas (Table 8.4), was avoidance of grasshopper mice. Alternatively, because insects are important prey for both deer mice and grasshopper mice during the spring, competition might explain habitat use of both species. In response to experimental removal of grasshopper mice, deer mice remained higher on removal plots than on controls (Stapp 1997a); declines in deer mouse abundance were negatively correlated with abundance of grasshopper mice and with the amount of shrub cover on study

150

Ecology of the Shortgrass Steppe

plots. Deer mice actually increased their use of shrubs when grasshopper mice were most abundant, suggesting that shrubs may provide some refuge from the more stocky-bodied grasshopper mice. Because consumption of insects by deer mice was unaffected by the removal experiment, and because granivorous rodents (kangaroo rats and harvest mice) also increased in abundance on removal plots, aggressive or predatory interference, rather than exploitative competition, was the most likely explanation (Stapp, 1997a). Predation by grasshopper mice is probably not an important source of mortality for adult deer mice and other rodents; however, high densities of grasshopper mice may affect local abundance of other rodents directly (through opportunistic predation on juveniles or litters in burrows) or indirectly (by affecting activity and habitat use). Overall, patterns of the local distribution and abundance of small mammals emerge from the interactions among species and their resources in the context of the range of environmental conditions in a given area (Fig. 8.6). For species such as deer mice, western harvest mice, and probably prairie voles, local population density is determined largely by the amount of vegetative cover, which provides both food and protection from predators. Granivorous rodents such as kangaroo rats and the less common pocket mice (Chaetodipus, Perognathus) respond primarily to soil type through its effect on the production and availability of palatable seeds. The distribution of grasshopper mice also reflects edaphic factors, but indirectly via the effects of soil friability on the density and activity of burrowing rodents, and as a consequence, the availability of arthropod prey. In the shrub-dominated areas of the shortgrass steppe, grasshopper mice may modify the behavior and population dynamics of deer mice and other rodents, although

SOIL PROPERTIES SHRUBS

GRASS/FORBS CACTUS GOPHER MOUNDS

cover

Reithrodontomys megalotis

seeds

Dipodomys ordii

arthropods

Peromyscus maniculatus

Onychomys leucogaster

Figure 8.6 Ecological relationships among habitat characteristics (uppercase), resources (lowercase), and interactions among the four most common nocturnal rodents (italics) in the northern shortgrass steppe. (Modified from Stapp [1996].)

Ecology of Mammals of the Shortgrass Steppe 151

these effects will likely vary seasonally with the availability of insect prey. Only grasshopper mice and diurnal ground squirrels are typically present in most shortgrass habitats, however, and only at very low densities (Table 8.4), which makes it logistically difficult to conduct intensive population studies. These circumstances suggest that a long-term, comparative approach is necessary to understand what factors limit the dynamics and distribution of mammal populations.

Long-Term Dynamics: Integrating IBP Grassland Biome and SGS LTER Studies Although more than a decade passed between the Grassland Biome and SGS LTER mammal studies, the fact that small mammals were the focus of research on the same site during both projects provides an opportunity to make some longterm comparisons. We compared temporal variation in the relative abundance of the two most common nocturnal rodents—deer mice and grasshopper mice (Fig. 8.7)—for a span of 35 years. As in other systems (e.g., Brown and Heske, 1990; Kesner and Linzey, 1997), Peromyscus populations tended to fluctuate from year to year, whereas grasshopper mouse numbers remained relatively constant over most of the study period. Deer mice are more generalized in their diet than grasshopper mice, which are larger, longer lived, and have a lower reproductive rate (McCarty, 1978). Fluctuations in abundance of generalist species such as deer

% of long-term mean

500

deer mouse northern grasshopper mouse

400 300 200 100 0 1970

1975

1985

1990

1995

2000

2005

Year Figure 8.7 Temporal changes in the relative abundance of deer mice and grasshopper mice on the Pawnee/SGS LTER site in north–central Colorado. Values are expressed as a percentage of the mean abundance from Grassland Biome studies (1971–1975 [Abramsky, 1976, French and Grant, 1974]), saltbush trapping grids (1987–1993 [L. McEwen, unpublished data]), and SGS LTER long-term monitoring studies (September trapping sessions in saltbush sites, 1994–2006 [P. Stapp, unpublished data]).

Ecology of the Shortgrass Steppe

Oct. - Sept. precipitation (mm)

152

600 500 400 300 200 100 0 1970

1975

1980

1985

1990

1995

2000

2005

Year Figure 8.8 Annual precipitation (measured in millimeters) expressed as totals from October through September, from the Grassland Biome and SGS LTER meteorological station (station 11). The dashed line represents the annual mean between October 1968 to September 2006.

mice may be associated with weather-related variation in food or cover; the 1997 population crash followed the wettest summer in 35 years of records for our study area (Fig. 8.8). The 1975 decline in grasshopper mouse numbers (Fig. 8.7) may have reflected higher plant cover on control plots of the nitrogen and water addition experiment, which had been removed from cattle grazing for 6 years and may have been affected by seed production of exotic species in the neighboring treatments. Deer mice and, to a lesser extent, grasshopper mice experienced marked declines in 2000 in association with the onset of drought conditions that persisted through at least 2004. We also compared our more recent estimates of rabbit density with those conducted during the Grassland Biome studies (Table 8.6). These results suggest that desert cottontails may be more abundant now than during the IBP studies, but that, until recently, jackrabbit densities were similar to those 30 years ago. Our recent surveys also suggest that white-tailed jackrabbits may be more abundant now than during the Grassland Biome studies (P. Stapp, unpublished data). Since approximately 2004, abundance of cottontails and black-tailed jackrabbits has increased dramatically (Fig. 8.5), which may reflect the combined effects of the exponential increase in density of prairie dogs on the CPER, and the period of drought from 2000 to 2004 (Fig. 8.8). Interestingly, J. Fitzgerald of UNC (December 1997) indicated that both cottontails and jackrabbits were much more abundant during the late 1970s, the last time when annual precipitation was less than 300 mm for multiple years. The nitrogen and water addition experiment conducted during the Grassland Biome studies showed that the addition of these resources to native shortgrass

Ecology of Mammals of the Shortgrass Steppe 153 Table 8.6 Population Densities of Rabbits in Northern Shortgrass Steppe Site Grassland Biome (1970–1971) Black-tailed jackrabbit White-tailed jackrabbit Desert cottontail SGS LTER (1994–1996) Black-tailed jackrabbit White-tailed jackrabbit Desert cottontail

Density, no./km 2

Total, %

5.86 1.63 2.66

57.73 16.06 26.21

6.04 1.23 6.42

44.12 8.98 46.90

Population densities of rabbits in the late 1990s in the northern shortgrass steppe were similar to those during the Grassland Biome studies of the early 1970s. Densities were calculated from Flinders and Hansen (1973, 1975) and from SGS LTER population monitoring studies (P. Stapp, 1996, unpublished data). Numbers of both cottontails and, especially, black-tailed jackrabbits, increased dramatically starting in 2004 (see Fig. 8.5).

plots had dramatic short-term effects on both plant and rodent communities (Grant et al., 1977; Lauenroth et al., 1978). Milchunas and Lauenroth (1995) investigated the long-term consequences of these treatments by continuing to sample vegetation on the experimental plots after the termination of the experiment in 1975. They found that the initial disturbance had far-reaching and unpredictable effects on plant community structure, driven primarily by time lags in the effects of litter accumulation in water and nitrogen-plus-water plots. Given these dynamic shifts in the plant community, we asked: How have resident rodent populations responded to changes in vegetation and habitat structure during the past two decades? In September 1995 and 1996, we trapped nocturnal small mammals on the original eight plots and compared abundance and species composition with results from the last years of the Grassland Biome experiment (1974–1975) (French and Grant, 1974; Grant et al., 1977). Recognizing that changes in abundance might reflect differences in habitat characteristics among treatments, we also collected data on vegetation (percentage cover, species richness, densities of shrubs, cacti, and exotics) and substrate (percentage bare soil, density and area of animal disturbances) on each plot. Recall that, during the U.S. IBP studies, changes in rodent populations paralleled shifts in plant biomass between “wet” (water, nitrogen-plus-water) and “dry” (nitrogen, control) treatments, with invasion by voles and harvest mice onto highly productive, but disturbed, water and nitrogen-plus-water plots (Grant et al., 1977). A cluster analysis on habitat variables measured in 1996 indicated that clear differences in vegetation and substrate characteristics between wet and dry treatments still persisted some 20 years later (Fig. 8.9). A cluster analysis using rodent densities from 1974 to 1975 produced a similar grouping of sites (P. Stapp, unpublished data). By 1995 to 1996, the rodent communities of these sites had changed dramatically (Fig. 8.10). In response to the long-term disturbance created by the treatments, both native and exotic species had dispersed among plots, some

154

Ecology of the Shortgrass Steppe

NW2 NW2 W1

Figure 8.9 Cluster analysis showing continuing similarity in vegetation and substrate in 1996 among replicate experimental plots where nitrogen (N), water (W), or both (NW) were added in 1971 as part of the Grassland Biome studies. C, controls. Wet (W, NW) and dry (N, C) treatments were still recognizably different 25 years after the experimental treatments began.

W2 C2 C1 N2 N1

0.00

0.25 0.50 0.75 1.00 Distance between groups

1.25

had gone locally extinct, and plots had been invaded by an additional three species of rodents (plains harvest mice, R. montanus; hispid pocket mice; and house mice) that were never captured during any of the Grassland Biome studies. These effects are in part confounded by the removal of grazing, because the pasture containing the plots had been fenced off from cattle in 1969 (Lauenroth et al., 1978). The extent of disturbance caused by long-term grazing exclusion is clear when one compares plant species composition and abundance on the control plots (C in Fig. 8.10) with those from outside the exclosure on native prairie (SG in Fig. 8.10) and with other long-term grazing exclosures (EX in Fig. 8.10). Curiously, kangaroo rats, which were captured previously on treatment plots (Abramsky, 1978; French and Grant, 1974), and which were frequently seen foraging along the adjacent gravel road during our trapping, were never captured by us on any of the experimental plots in 1995 to 1996. We conclude from these results that long-term responses to disturbance of both plants and small mammals of the shortgrass steppe are highly dynamic, persistent, and unpredictable. Even though the effects of treatments were still detectable 25 years later, in the absence of grazing, both manipulated and control plots clearly were disturbed compared with the adjacent grasslands. Given the small size and close proximity of the study plots, and in light of increasing recognition of the movement abilities of small mammals (e.g., Stapp, 1997b), we speculate that by 1995, rodents may have perceived habitat features of all plots in the pasture as similar, regardless of the initial treatment. Although some patterns from the earlier studies remain (e.g., more exotic plant species on nitrogen-plus-water plots), the apparent differences in rodent abundance and species composition among plots may best reflect random variation in the location and capture rates of individuals interacting with habitat at a larger scale. Future studies should take into account the relative homogeneity of shortgrass steppe vegetation for most species and the overwhelming role that less common habitats such as shrublands, road margins, and fence lines might play in dynamics at the landscape scale.

Ecology of Mammals of the Shortgrass Steppe 155

60 DIOR CHHI MUMU MIOC REMO REME ONLE PEMA

Density (# ha-1)

55 50 15

10

5

0

C

N W NW

1974-75

C

N W NW SG EX

1995-96 Treatment

Figure 8.10 Comparisons between population density and species composition of nocturnal small mammals during and 20 years after the end of nutrient stress experiments to examine the effects of nitrogen (N), water (W), and nitrogen and water combined (NW), on plant and small-mammal communities of the shortgrass steppe. C, controls. Values are means from two replicate plots of each treatment. Density estimates from two nearby upland shortgrass prairie plots (SG) and from two long-term grazing exclosures trapped in 1996 (EX) are provided for comparison. Species abbreviations: CHHI, hispid pocket mouse; DIOR, Ord’s kangaroo rat; MIOC, prairie vole; MUMU, house mouse; ONLE, northern grasshopper mouse; PEMA, deer mouse; REME, Western harvest mouse; REMO, plains harvest mouse.

Role of Mammals in the Shortgrass Steppe Ecosystem Much of research during the Grassland Biome and SGS LTER projects was focused on understanding the role of mammals in grassland ecosystems. These studies have demonstrated that mammals perform two central functions that have influenced the evolution and ecology of the shortgrass steppe: removal of primary productivity by herbivore grazing, and burrow and mound construction by small mammals. Herbivory Larson (1940) was one of the first to argue that grazing by native herbivores such as bison was integral in the evolution and maintenance of the shortgrass steppe, citing historical records of expansive shortgrass plains populations prior to widespread introduction of livestock and the relative tolerance of the dominant

156

Ecology of the Shortgrass Steppe

perennial grasses to overgrazing. Although wild bison were extirpated in the shortgrass steppe during the early 20th century (Fitzgerald et al., 1994), low to moderate grazing by cattle may perform a similar function in maintaining plant communities in the shortgrass steppe (Milchunas et al., 1989). Hart and Derner (chapter 17, this volume) and Milchunas et al. (chapters 16 and 18, this volume) provide excellent reviews of the history and effects of herbivory by mammals, especially native ungulates and livestock, so our discussion here will be brief and limited to the general effects of small mammals on ecosystems. Unlike more productive grasslands, small rodent herbivores such as voles and cotton rats are relatively uncommon in the shortgrass steppe, and therefore have minimal impacts on vegetation. Prairie dogs typically occupy lowlying areas with relatively fine-textured soils (Table 8.3), where sedges and buffalograss predominate, but they can affect vegetation structure and plant species composition in these areas by selectively consuming certain plant species and by providing sites of establishment for forbs and exotic species (Barko et al., 1999; Bonham and Lerwick, 1976; Hartley, 2006; Klatt and Hein, 1978; Severe, 1977; Stapp, 2007; Winter et al., 2002). Comparatively little is known of the effects of browsing by rabbits, which are the dominant small herbivores in terms of biomass, on grasses and shrubs (Lauenroth and Milchunas, 1991). Both jackrabbit species prefer western wheatgrass (A. smithii), and Flinders and Hansen (1972) suggested that consumption by jackrabbits may significantly affect the vegetation in their habitats. Similarly, in coarsely textured soils and in disturbed areas, selective storage and consumption of seeds by granivores can influence the distribution of plants. Working in abandoned agricultural plots on the CPER, Hoffman (1992; Hoffman et al., 1995) showed that seed predation by kangaroo rats and pocket mice significantly affected seedling establishment of large-seeded grass species. Grant and French (1980) used simulation models to evaluate the role of small mammals in grassland ecosystems. They concluded that, even under high population densities, direct consumption by small mammals has an insignificant effect on primary production. This is particularly true for the shortgrass steppe, where herbivorous rodents (Microtus, Sigmodon) are absent from most habitats. Their simulations suggested that small mammals can affect other ecosystem components in two ways. First, consumption may have a significant effect on aboveground arthropod biomass dynamics, which may be particularly significant in the shortgrass steppe, with fauna and biomass that are dominated by omnivores and insectivores (Table 8.4). Second, the upward translocation of soil to the ground surface caused by the burrowing of small mammals significantly alters the percentage of bare surface soil and, especially, the amount and rate of processing of soil organic matter to make nutrients available to plants. The effects of fossorial and semifossorial rodents on the flora, and especially the fauna of the shortgrass steppe are described in the next section. Burrowing and Mound Construction The most successful mammals in the shortgrass steppe construct burrows to minimize exposure to inclement weather and predators or as a by-product of their

Ecology of Mammals of the Shortgrass Steppe 157

foraging activities. In the process, they transport shallow soil organic matter and mineral soils to the surface, increase soil erosion and filtration rates, and alter surface microclimates (Koford, 1958). At normal densities, small mammals introduce an estimated 5 to 6 kg nitrogen (N)⋅ha–1⋅y–1 to the top soil layer, which is similar to or greater than that contributed by precipitation and other sources (3–5 kg N⋅ha–1⋅y–1) (Grant and French, 1980). These benefits are balanced by the detrimental effects of digging and mound construction on the mortality of individual plants (Peters et al., chapter 6, this volume). In some cases, mounds also serve as seedbank and establishment sites for less common herbaceous and grass species that otherwise could not compete with B. gracilis and B. dactyloides. More important, burrows and mounds provide critical refuge habitats for a variety of grassland animals. The effects of burrowing mammals on grassland ecosystem function and diversity differ among species according to their tolerances for different soil and topographic conditions, to the distribution of preferred food plants, and to their population densities. Swift foxes, badgers, and coyotes all construct and maintain holes (e.g., Egoscue, 1979; Linzey, 1982) that can have significant effects (Platt, 1975), but in the shortgrass steppe, small mammals are the most significant source of burrows. Of these, prairie dogs (Cynomys) and pocket gophers (Thomomys, Geomys, Cratogeomys) are probably the most important because of the intensity and persistence of disturbances they create. Prairie Dogs It is widely recognized that prairie dogs (Fig. 8.11), through their grazing and burrowing activities, significantly alter structure and function of grassland ecosystems

Figure 8.11 Black-tailed prairie dogs play an important and often controversial role in the ecology of many areas of the shortgrass steppe. (Photo by Stephen J. Dinsmore.)

158

Ecology of the Shortgrass Steppe

in areas where soil and topographic conditions favor the establishment of colonies (Whicker and Detling, 1988). Prairie dog burrows also provide shelter and nesting locations for a variety of other grassland animals (Stapp, 1998), including birds such as Burrowing Owls (e.g., Butts and Lewis, 1982; Desmond et al., 2000; Orth and Kennedy, 2001; Sidle et al., 2001a) and Mountain Plovers (Charadrius montanus [Dinsmore et al., 2005; Dreitz et al., 2005]), both species of conservation concern in the western Great Plains. Prairie dogs are the primary prey of the endangered black-footed ferret, which faces extinction largely because of the widespread extirpation of prairie dogs (Anderson et al., 1986; Calahane, 1954). Prairie dogs are also an important food source for many grassland raptors (Cully, 1991; Plumpton and Andersen, 1997, 1998; Schmutz and Fyfe, 1987; Seery and Matiatos, 2000; Weber, 2001). The recent increase in shortgrass steppe populations of Ferruginous and Swainson’s Hawks (Wiens and McIntyre, chapter 9, this volume) may, in part, reflect the recovery of prairie dog colonies in eastern Colorado during the past two decades. Despite their ecological importance, prairie dogs are often considered rangeland pests because of their effects on production and availability of forage for livestock (Derner et al., 2006; Hansen and Gold, 1977; O’Meilia et al., 1982). Prairie dogs still occupy most of their historical geographic range, but local populations have been eliminated or reduced by extensive poisoning and shooting campaigns associated with ranching and farming, and by sylvatic plague (Miller et al., 1990). Existing colonies are thought to cover less than 10% of the 41 million ha occupied in 1900 (Anderson et al., 1986), although there is some debate about the historical coverage of prairie dog colonies (Forrest, 2005; Knowles et al., 2002; Vermeire et al., 2004; Virchow and Hygnstrom, 2002). Although blacktailed prairie dogs occur throughout the shortgrass steppe, populations appear to be relatively small and disjunct compared with historical accounts. On the PNG in northeastern Colorado, most (67%) colonies are less than 20 ha in size, and the total area occupied there by prairie dogs in 2006 (1149 ha) represented only about 1.5% of the total area (78,000 ha) (USDA Forest Service, unpublished data). This coverage seems typical for colonies of black-tailed prairie dogs elsewhere (Butts and Lewis, 1982; Clark et al., 1982; Sidle et al., 2001b), although larger colonies occurred historically. Because populations are naturally fragmented by variation in topography, soils, and abundance of preferred plants (Koford, 1958), it seems likely that colonies never covered more than 10% to 20% of the landscape. However, through their burrowing and grazing activities and by serving as prey, prairie dogs may have a disproportionately large impact on both shortgrass steppe vegetation and native fauna, at both local and landscape scales. Between 1997 and 1999, we conducted comparative studies to determine the ecological role of prairie dogs in the northern shortgrass steppe. Plots (1.2 ha) were established on five different prairie dog colonies and similar control plots on neighboring areas with the same soil type, topography, vegetation, and landuse history, but without prairie dogs. We hypothesized that the effects of prairie dogs on the shortgrass steppe differed from those on mixed-grass prairie in South Dakota and Montana, where much of the earlier research had been conducted. In these more productive grasslands, the activities of prairie dogs create patches

Ecology of Mammals of the Shortgrass Steppe 159

of low grassland in a sea of taller vegetation, creating greater landscape-scale heterogeneity. By comparison, in most areas of the shortgrass steppe, the contrast between the vegetation of colonies and that of native prairie would be less obvious to us and to the resident fauna. However, because of the scarcity of belowground refuges and low consumer densities on the shortgrass steppe, prairie dog burrows provide critical habitat for other animals, and prairie dogs themselves are important prey for top vertebrate predators (Stapp, 1998). Our comparative studies suggest that plant communities and fauna of prairie dog colonies differ in several ways from grasslands where prairie dogs are absent (Table 8.7). As found in other studies in the shortgrass steppe (Bonham and Lerwick, 1976; Severe, 1977), burrowing by prairie dogs seems to alter plant species composition in the area of the mounds by providing germination sites for exotic and other species associated with disturbance. Plant species associated exclusively with mounds included Cleome serrulata, Solanum triflorum, Portulaca oleracea, Chenopodium album, Euphorbia glyptosperma, and Salsola iberica. Unlike Bonham and Lerwick’s (1976) results, however, there were no consistently significant differences in plant species richness between colonies and uncolonized grasslands. Grazing and burrowing by prairie dogs produced slight but significant changes in vegetation height, and an increase in the percentage cover and patchiness of bare soil (Bonham and Hannan, 1978), but did not significantly alter the microtopography relative to the controls (Table 8.7). These results are similar to those reported by Stapp (2007) in a comparison of 18 active colonies with seven uncolonized control sites on the PNG, and are generally consistent with other studies of vegetation responses to prairie dogs in the southern shortgrass steppe (Barko et al., 2001; Winter et al., 2002). Except for grasshoppers, which tended to be more abundant outside of prairie dog colonies than inside colonies (J. R. Junell and B. Van Horne, unpublished data), changes in habitat structure apparently had little effect on abundance of arthropods or most of their vertebrate predators (Table 8.7) (Kretzer and Cully, 2001b). There was one exception, in that lesser earless lizards (Holbrookia maculata) were more abundant on prairie dog colonies than off. However, there were no striking differences in abundance of other amphibian or reptile species (Davis and Theimer, 2003; Kretzer and Cully, 2001a). Rodent burrows, including those of prairie dogs, are known to serve as primary overwintering sites for western rattlesnakes (Crotalus viridis) and other reptiles (Hammerson, 1999), but we were not able to confirm that, because our field surveys were limited to the warmer months (May–September). Horned Larks (Eremophila alpestris), a bird species of open grasslands, were more common on prairie dog colonies than on adjacent uncolonized grassland plots, whereas Lark Buntings (Calamospiza melanocorys) were significantly less abundant on prairie dog colonies than off (Table 8.7). Avian species richness was significantly higher on prairie dog colonies, presumably as a result of raptors preying on prairie dogs (M. Andre and P. Stapp, unpublished data). Similarly, Barko et al. (1999) recently reported higher avifaunal abundance and diversity in five Oklahoma prairie dog colonies than in grasslands without prairie dogs, but only during the plant growing season (Winter et al., 2003). Smith and Lomolino (2004)

160

Ecology of the Shortgrass Steppe

Table 8.7 Ecological Effects of Prairie Dogs in the Northern Shortgrass Steppe Based on SGS LTER Comparative Studies from Four to Five Paired Colony and Grassland Control Plots in 1997 and 1998 Group Plants and cover Mound scale

Colony scale

Arthropods

Vertebrates Amphibians and lizardsb

Birdsc

Mammals

Colonies (vs. Grassland Controls) More exotic and disturbance species on mounds; higher density of Sphaeralcea coccinea; lower density of shortgrass species, Plantago patagonica, and lichens; no difference in plant species richness or the number of unique species compared with off mounds No difference in plant species richness; shorter grasses and shrubs; greater percentage cover and higher spatial variation of bare soil; no difference in microtopographic variation No difference in relative abundance of beetles, crickets, or spiders; fewer grasshoppersa; no difference in density of harvester ant mounds No difference in species richness; no differences in abundance of western chorus frogs, tiger salamanders, or short-horned lizards; higher numbers of lesser earless lizards; fewer northern manylined skinks Higher species richness (resulting from the presence of raptors); horned larks, burrowing owls, raptors more abundant; fewer lark buntings; no differences in abundance of western meadowlarks, McCown’s longspurs, or all birds combined; lower densities of passerine nests and lower survival rates of artificial nests Higher small-mammal species richness; higher numbers of northern grasshopper mice and Ord’s kangaroo rats; fewer thirteenlined ground squirrels; higher density of cottontails and coyote scats; no difference in density of fox and badger burrows; no difference in cattle activity (density of fecal pats)

a

Junell, 2002. α ≤ 0.20 because of low capture rates and, therefore, low power. c M. Andre and P. Stapp, (unpublished data). Unless noted, all results were statistically significant at P ≤ .10 using paired t-tests. Species not mentioned were captured or sighted too infrequently to compare abundance statistically (P. Stapp, unpublished data). b

also found strong seasonal differences in bird communities between colonies and other vegetation types in Oklahoma, with Horned Larks showing the most consistent differences in abundance on and off colonies. In their study, species richness and community structure in summer were highest on prairie dog colonies and fallow crop fields, but most differences among vegetation types were reduced during fall. In addition to comparing avian population densities and diversity on and off prairie dog colonies, we conducted artificial nest experiments on paired plots in 1998. Densities of passerine nests were higher, and rates of nest predation were lower, on grassland control plots than on prairie dog colonies (M. Andre and P. Stapp, unpublished data). Baker et al. (1999, 2000) similarly found higher rates of predation of artificial nests on colonies of white-tailed and black-tailed prairie dogs, respectively, than in adjacent grasslands lacking prairie dogs.

Ecology of Mammals of the Shortgrass Steppe 161

Prairie dogs may also have significant effects on other small mammals. Northern grasshopper mice were more abundant in colonies and, on average, more species of small mammals were captured in colonies than in nearby grasslands (Table 8.7). In contrast, thirteen-lined ground squirrels were less abundant in prairie dog colonies. A more comprehensive study of small-mammal communities in active and inactive colonies (Stapp, 2007) similarly found higher numbers of grasshopper mice in colonies than in uncolonized grasslands, and lower numbers of ground squirrels in active colonies than in inactive ones, but no difference in overall species richness or community patterns. The lack of strong differences in rodent communities on and off colonies is consistent with other studies throughout the region, with local vegetation and habitat characteristics likely playing more important roles (Stapp, 2007). The potential for interactions between prairie dogs and ground squirrels, however, warrants additional attention. There was no evidence from microhabitat characteristics at trapping locations that ground squirrels avoided areas of high prairie dog activity (P. Stapp, unpublished data); however, vigilance behavior of juvenile ground squirrels was significantly higher on active prairie dog colonies than on colonies that had been recently extirpated by plague (J. Aldana and P. Stapp, unpublished data). There may be subtle agonistic interactions between prairie dogs and ground squirrels, or prairie dogs may affect food availability or habitat quality for ground squirrels, or they may attract predators to colonies that subsequently prey on squirrels. We spotted significantly higher numbers of desert cottontails and scats of coyotes on prairie dog colonies than on grassland control plots (Table 8.7). Dano (1952) and Hansen and Gold (1977) also found higher densities of rabbits in prairie dog colonies, which may provide an additional source of prey for both avian and mammalian predators, including coyotes. Shaughnessy and Cifelli (2004) reported few differences in carnivore abundance on and off prairie dog colonies in Oklahoma, with badgers (T. taxus) and spotted skunks (Spilogale putorius) more common on colonies in Cimarron County but not in the Oklahoma Panhandle (Lomolino and Smith, 2003). Finally, other researchers have suggested that cattle and other ungulates concentrate activity on prairie dog colonies (Coppock et al., 1983; Knowles, 1986; Koford, 1958; Krueger, 1986; Lomolino and Smith 2003), but in our studies, density of fecal pats on prairie dog colonies were similar to those on our control plots (Table 8.7) and to plots in low and moderately grazed pastures (P. Stapp, unpublished data). Guenther and Detling (2003) also reported that cattle neither preferred nor avoided prairie dog colonies at the SGS LTER site. Collectively, our results to date suggest that prairie dogs may have a significant influence on the flora and fauna of the shortgrass steppe, although the effects of prairie dogs are less striking than reported in studies in more productive grasslands, where the contrast between colonies and surrounding grasslands is more pronounced. Animals that benefit most from the presence of prairie dogs include avian and mammalian predators, species associated with heavy grazing and/or low stature grasslands (e.g., Mountain Plovers, Horned Larks, lesser earless lizards), and species that depend on abandoned burrows for shelter (e.g., Burrowing Owls, northern grasshopper mice, desert cottontails, skunks). Except for a few notable species of conservation concern, all the species recorded in prairie dog colonies

162

Ecology of the Shortgrass Steppe

are regularly observed in areas without prairie dogs, suggesting that effects of prairie dogs are primarily reflected in changes in relative abundance of the most common consumers. These results are not surprising, given the relatively small amount of our study area occupied by prairie dogs and the relatively low population densities of many animals in the shortgrass steppe. However, combined with the effects of high densities of prairie dogs and their effects on ecosystem functioning, they highlight the importance of prairie dog colonies for increasing grassland biodiversity at larger, landscape scales. Because prairie dog colonies in the shortgrass steppe tend to be small and somewhat isolated, it is critical to understand the biology and dynamics of prairie dogs to evaluate the ecological consequences of their historical declines. Roach et al. (2001) used microsatellite markers to describe the genetic structure of SGS LTER prairie dog populations. The 13 colonies they studied ranged greatly in active area (1–52 km2) and age (1–10 years), in part as a result of recent extinctions caused by plague and human eradication efforts. Roach et al. (2001) reported a moderate amount of genetic relatedness among colonies, suggesting some degree of isolation and differentiation, but they found little evidence of inbreeding within colonies. They argued that, despite the distance separating colonies, the genetic evidence indicated a considerable movement of individuals among populations, especially among large, persistent colonies. Colony area and age were highly correlated, and colony size/age was the best predictor of the degree of genetic relatedness among colonies. Small, recently established colonies tended to differ more from one another than larger, more permanent ones, which suggests a mainland–island metapopulation structure, in which small satellite sites with high extinction rates are recolonized by dispersers from large colonies. Dispersal was facilitated by a network of swales and ephemeral drainages, areas where preferred food plants (B. gracilis, Carex eleocharis, B. dactyloides, A. smitthii) (Bonham and Lerwick, 1976; Hansen and Gold, 1977; Koford, 1958) are abundant and where most colonies are located. The physical distance between colonies via drainages explained a small (60 ha) parcels of native shortgrass steppe against conversion to cropland or urban areas (Boren et al., 1997; Walk and Warner, 1999). • Reclaim former grassland areas (e.g., through the CRP). • Preserve a variety of macrohabitat types (e.g., riparian areas, shortgrass uplands) within these landscapes. • Focus conservation and management actions on functioning landscapes, not just isolated parcels of grassland. • Recognize the value of grazing as a bona fide, ecologically sound management practice, and maintain both grazed and ungrazed areas. • Develop a broader understanding of the factors and areas that limit grassland bird populations. • Develop networks of linked conservation and management areas to enhance connectivity and promote conservation of wintering and breeding grounds outside the shortgrass steppe (e.g., The Nature Conservancy’s Prairie Wings Program [www.nature.org/initiatives/programs/birds/explore/]). • Support legislative protection of rare, threatened, and endangered bird species of the shortgrass steppe. • Promote continued monitoring of bird diversity and density on the shortgrass steppe.

Birds of the Shortgrass Steppe 201

Clearly, we need to know more about the mechanisms that determine population fluctuations (particularly for declining species), differences in habitat occupation, and the sensitivity to changes in habitat structure and heterogeneity that may be produced by changing land use and climate change. To place the findings of such studies on a mechanistic foundation, studies are also needed of long-term reproductive success and survivorship under various grazing regimes. The research tradition of the NSF’s LTER program should foster continued ornithological studies at the SGS LTER site and elsewhere (see Collins, 2001). Our limited understanding of avian community and population patterns and the mechanisms accounting for these patterns has come from years of effort by numerous biologists and birders. Further advances in our ornithological knowledge will undoubtedly take a similarly strenuous effort, as the birds of the shortgrass steppe hold their secrets tightly.

Concluding Remarks In the end, it may miss the point to ask what birds do in ecosystems. Perhaps the notion that all species must have some function in the ecosystem reflects an overly enthusiastic application of a holistic philosophy of nature, the conviction that everything in nature has a purpose. But the challenge may go more deeply. Tight structuring in variable systems such as grasslands may simply not be possible, and highly vagile consumers such as birds may live, in a sense, off the variance in production, without offering a large regulatory or functional role in return. Perhaps, after all, the role of birds is to embody the special magic of grasslands, not as systems to be dissected and modeled, but as landscapes to be experienced and cherished.

Acknowledgments We are indebted to Ron Ryder for providing us with species lists and population data. Chris Helzer graciously provided information on avian responses to grassland fragmentation from his Masters of Science research. We thank Jim Miller, Susan Skagen, Paul Stapp, and Peter Vickery for critiquing manuscript drafts; Rich Strauss for statistical advice; and Fritz Knopf for discussions on Mountain Plovers.

Appendix: Checklist of Bird Species Occurring at the CPER Site, 1962–2005 Key: A, abundant; C, common; M, species present only as migrants; R, rare species; S, species that breed or are present in summer; U, uncommon; W, species that overwinter; Y, year-round residents. Terminology follows National Audubon Society usage and standards.

202

Ecology of the Shortgrass Steppe

Gaviiformes Common Loon (Gavia immer)

W

R

S W M M M

U R U U R

American White Pelican (Pelecanus erythrorhynchos) Double-crested Cormorant (Phalacrocorax auritus)

S S

U U

Ciconiiformes American Bittern (Botaurus lentiginosus) Great Blue Heron (Ardea herodias) Snowy Egret (Egretta thula) Green Heron (Butorides virescens) Black-crowned Night Heron (Nycticorax nycticorax) Yellow-crowned Night Heron (Nycticorax violacea) White-faced Ibis (Plegadis chihi)

M S S S S S M

R U R R U R U

M M M M W Y Y M Y M S S S Y M M S M M M M M M M S

R U R R U U R C C C C C U C U R U U R C C C C R U

S M

U R

Podicipediformes Pied-billed Grebe (Podilymbus podiceps) Horned Grebe (Podiceps auritus) Eared Grebe (Podiceps nigricollis) Western Grebe (Aechmophorus occidentalis) Clark’s Grebe (Aechmophorus clarkii) Pelecaniformes

Anseriformes Greater White-fronted Goose (Anser albifrons) Snow Goose (Chen caerulescens) Ross’ Goose (Chen rossii) Brant (Branta bernicla) Cackling Goose (Branta hutchinsii) Canada Goose (Branta canadensis) Wood Duck (Aix sponsa) Green-winged Teal (Anas crecca) Mallard (Anas platyrhynchos) Northern Pintail (Anas acuta) Blue-winged Teal (Anas discors) Cinnamon Teal (Anas cyanoptera) Northern Shoveler (Anas clypeata) Gadwall (Anas strepera) American Wigeon (Anas americana) Canvasback (Aythya valisineria) Redhead (Aythya americana) Ring-necked Duck (Aythya collaris) Greater Scaup (Aythya marila) Lesser Scaup (Aythya affi nis) Common Goldeneye (Bucephala clangula) Bufflehead (Bucephala albeola) Common Merganser (Mergus merganser) Red-breasted Merganser (Mergus serrator) Ruddy Duck (Oxyura jamaicensis) Falconiformes Turkey Vulture (Cathartes aura) Osprey (Pandion haliaetus)

Birds of the Shortgrass Steppe 203 Mississippi Kite (Ictinia mississippiensis) Bald Eagle (Haliaeetus leucocephalus) Northern Harrier (Circus cyaneus) Sharp-shinned Hawk (Accipiter striatus) Cooper’s Hawk (Accipiter cooperii) Northern Goshawk (Accipiter gentilis) Broad-winged Hawk (Buteo platypterus) Red-tailed Hawk (Buteo jamaicensis) Krider’s race Harlan’s race Swainson’s Hawk (Buteo swainsoni) Rough-legged Hawk (Buteo lagopus) Ferruginous Hawk (Buteo regalis) Golden Eagle (Aquila chrysaetos) American Kestrel (Falco sparverius) Merlin (Falco columbarius) Peregrine Falcon (Falco peregrinus) Gyrfalcon (Falco rusticolus) Prairie Falcon (Falco mexicanus)

M W Y Y M M M

R R C U U R R

Y W S W Y Y Y W M W Y

C R C C C C C C R R C

Y Y Y

U R U

M M Y M

U U C C

M M Y S S M M M M S S M S M M M M M M M

R R C C C C C U C C R R U R C R U U C R

Galliformes Ring-necked Pheasant (Phasianus colchicus) Sharp-tailed Grouse (Tympahuchus phasianellus) Northern Bobwhite (Colinus virginianus) Gruiformes Virginia Rail (Rallus limicola) Sora (Porzana carolina) American Coot (Fulica americana) Sandhill Crane (Grus canadensis) Charadriiformes Snowy Plover (Charadrius alexandrinus) Semipalmated Plover (Charadrius semipalmatus) Killdeer (Charadrius vociferus) Mountain Plover (Charadrius montanus) American Avocet (Recurvirostra americana) Greater Yellowlegs (Tringa melanoleuca) Lesser Yellowlegs (Tringa flavipes) Solitary Sandpiper (Tringa solitaria) Willet (Tringa semipalmata) Spotted Sandpiper (Actitis macularius) Upland Sandpiper (Bartramia longicauda) Whimbrel (Numenius phaeopus) Long-billed Curlew (Numenius americanus) Hudsonian Godwit (Limosa haemastica) Marbled Godwit (Limosa fedoa) Sanderling (Calidris alba) Semipalmated Sandpiper (Calidris pusilla) Western Sandpiper (Calidris mauri) Least Sandpiper (Calidris minutilla) White-rumped Sandpiper (Calidris fuscicollis)

204

Ecology of the Shortgrass Steppe

Baird’s Sandpiper (Calidris bairdii) Pectoral Sandpiper (Calidris melanotos) Stilt Sandpiper (Calidris himantopus) Long-billed Dowitcher (Limnodromus scolopaceus) Wilson’s Snipe (Gallinago delicata) Wilson’s Phalarope (Phalaropus tricolor) Red-necked Phalarope (Phalaropus lobatus) Franklin’s Gull (Larus pipixcan) Ring-billed Gull (Larus delawarensis) California Gull (Larus californicus) Forster’s Tern (Sterna forsteri) Least Tern (Sterna antillarum) Black Tern (Chlidonias niger)

M M M M Y S M M Y S S M S

C U R C U C R A C C U R C

Y M Y

C R C

S S

R U

S Y Y W S Y W W

U U C R C U R R

S S

C R

S S M S M

R U R U R

Y

U

M

R

Columbiformes Rock Dove (Columba livia) White-winged Dove (Zenaida asiatica) Mourning Dove (Zenaida macroura) Cuculiformes Black-billed Cuckoo (Coccyzus erythropthalmus) Yellow-billed Cuckoo (Coccyzus americanus) Strigiformes Common Barn-owl (Tyto alba) Eastern Screech-owl (Otus asio) Great Horned Owl (Bubo virginianus) Snowy Owl (Nyctea scandiacus) Burrowing Owl (Athene cunicularia) Long-eared Owl (Asio otus) Short-eared Owl (Asio fl ammeus) Northern Saw-whet Owl (Aegolius acadicus) Caprimulgiformes Common Nighthawk (Chordeiles minor) Common Poorwill (Phalaenoptilus nuttallii) Apodiformes Chimney Swift (Chaetura peliagica) White-throated Swift (Aeronautes saxatalis) Calliope Hummingbird (Stellula calliope) Broad-tailed Hummingbird (Selasphorus platycercus) Rufous Hummingbird (Selasphorus rufus) Coraciiformes Belted Kingfisher (Ceryle alcyon) Piciformes Lewis’ Woodpecker (Melanerpes lewis)

Birds of the Shortgrass Steppe 205 Red-headed Woodpecker (Melanerpes erythrocephalus) Yellow-bellied Sapsucker (Sphyrapicus varius) Red-naped Sapsucker (Sphyrapicus nuchalis) Williamson’s Sapsucker (Sphyrapicus thyroideus) Ladder-backed Woodpecker (Picoides scalaris) Downy Woodpecker (Picoides pubescens) Hairy Woodpecker (Picoides villosus) Northern Flicker (Colaptes auratus) Red-shafted race Yellow-shafted race

S W M M M Y W

U R U R R C U

Y Y

C U

S S S M S M S S M M M S S Y M M M S S S S Y M W Y Y W Y W W Y W S S M W M M S M M M

R C C R C R R R U U U C R A R U U C C C C C R R C C R C R C U C C C R U C C U R C U

Passeriformes Cassin’s Kingbird (Tyrannus vociferans) Western Kingbird (Tyrannus verticalis) Eastern Kingbird (Tyrannus tyrannus) Scissor-tailed Flycatcher (Tyrannus forficatus) Western Wood-pewee (Contopus sordidulus) Olive-sided Flycatcher (Contopus cooperi) Willow Flycatcher (Empidonax traillii) Least Flycatcher (Empidonax minimus) Hammond’s Flycatcher (Empidonax hammondii) Dusky Flycatcher (Empidonax oberholseri) Cordilleran Flycatcher (Empidonax occidentalis) Say’s Phoebe (Sayornis saya) Ash-throated Flycatcher (Myiarchus cinerascens) Horned Lark (Eremophilia alpestris) Purple Martin (Progne subis) Tree Swallow (Tachycineta bicolor) Violet-green Swallow (Tachycineta thalassina) Northern Rough-winged Swallow (Stelgidopteryx serripennis) Bank Swallow (Riparia riparia) Barn Swallow (Hirundo rustica) Cliff Swallow (Hirundo pyrrhonota) Blue Jay (Cyanocitta cristata) Pinyon Jay (Gymnorhinus cyanocephalus) Clark’s Nutcracker (Nucifraga columbiana) Black-billed Magpie (Pica hudsonia) American Crow (Corvus brachyrhynchos) Common Raven (Corvus corax) Black-capped Chickadee (Parus atricapillus) Mountain Chickadee (Parus gambeli) Red-breasted Nuthatch (Sitta canadensis) White-breasted Nuthatch (Sitta carolinensis) Brown Creeper (Certhia americana) House Wren (Troglodytes aedon) Rock Wren (Salpinctes obsoletus) Bewick’s Wren (Thryomanes bewickii) Golden-crowned Kinglet (Regulus satrapa) Ruby-crowned Kinglet (Regulus calendula) Blue-gray Gnatcatcher (Polioptila caerulea) Eastern Bluebird (Sialia sialis) Western Bluebird (Sialia mexicana) Mountain Bluebird (Sialia currucoides) Townsend’s Solitaire (Myadestes townsendi)

206

Ecology of the Shortgrass Steppe

Veery (Catharus fuscescens) Swainson’s Thrush (Catharus ustulatus) Hermit Thrush (Catharus guttatus) American Robin (Turdus migratorius) American Pipit (Anthus rubescens) Sprague’s Pipit (Anthus spragueii) Gray Catbird (Dumetella carolinensis) Northern Mockingbird (Mimus polyglottos) Sage Thrasher (Oreoscoptes montanus) Brown Thrasher (Toxostoma rufum) Bohemian Waxwing (Bombycilla garrulus) Cedar Waxwing (Bombycilla cedrorum) Northern Shrike (Lanius excubitor) Loggerhead Shrike (Lanius ludovicianus) European Starling (Sturnus vulgaris) White-eyed Vireo (Vireo griseus) Plumbeous Vireo (Vireo plumbeus) Yellow-throated Vireo (Vireo flavifrons) Warbling Vireo (Vireo gilvus) Red-eyed Vireo (Vireo olivaceus) Golden-winged Warbler (Vermivora chrysoptera) Tennessee Warbler (Vermivora peregrina) Orange-crowned Warbler (Vermivora celata) Nashville Warbler (Vermivora ruficapilla) Virginia’s Warbler (Vermivora virginiae) Yellow Warbler (Dendroica petechia) Chestnut-sided Warbler (Dendroica pensylvanica) Magnolia Warbler (Dendroica magnolia) Cape May Warbler (Dendroica tigrina) Black-throated Blue Warbler (Dendroica caerulescens) Yellow-rumped Warbler (Dendroica coronata) Myrtle’s race Audubon’s race Townsend’s Warbler (Dendroica townsendi) Black-throated Green Warbler (Dendroica virens) Blackburnian Warbler (Dendroica fusca) Prairie Warbler (Dendroica discolor) Palm Warbler (Dendroica palmarum) Bay-breasted Warbler (Dendroica castanea) Blackpoll Warbler (Dendroica striata) Black-and-white Warbler (Mniotilta varia) American Redstart (Setophaga ruticilla) Prothonotary Warbler (Protonotaria citrea) Worm-eating Warbler (Helmintheros vermivorum) Ovenbird (Seiurus aurocapillus) Northern Waterthrush (Seiurus noveboracensis) Kentucky Warbler (Oporornis formosus) MacGillivray’s Warbler (Oporornis tolmiei) Common Yellowthroat (Geothlypis trichas) Hooded Warbler (Wilsonia citrina) Wilson’s Warbler (Wilsonia pusilla) Canada Warbler (Wilsonia canadensis) Yellow-breasted Chat (Icteria virens) Summer Tanager (Piranga rubra) Scarlet Tanager (Piranga olivacea)

M M M Y W M S S S S W Y W S Y S M M M S M M M M M S M M M M

U C U C R R U C C C R U C C C R R R C R R R U R U C R R R R

M M M M M M M M M M M M M M M M M S M M M S M M

C C U R R R R R R R U R R R R R C C R A R U R R

Birds of the Shortgrass Steppe 207 Western Tanager (Piranga ludoviciana) Rose-breasted Grosbeak (Pheucticus ludovicianus) Black-headed Grosbeak (Pheucticus melanocephalus) Blue Grosbeak (Passerina caerulea) Lazuli Bunting (Passerina amoena) Dickcissel (Spiza americana) Green-tailed Towhee (Pipilo chlorurus) Spotted Towhee (Pipilo maculatus) Cassin’s Sparrow (Aimophila cassinii) American Tree Sparrow (Spizella arborea) Chipping Sparrow (Spizella passerina) Clay-colored Sparrow (Spizella pallida) Brewer’s Sparrow (Spizella breweri) Vesper Sparrow (Pooecestes gramineus) Lark Sparrow (Chondestes grammacus) Black-throated Sparrow (Amphispiza bilineata) Sage Sparrow (Amphispiza belli) Lark Bunting (Calamospiza melanocorys) Savannah Sparrow (Passerculus sandwichensis) Baird’s Sparrow (Ammodramus bairdii) Grasshopper Sparrow (Ammodramus savannarum) Fox Sparrow (Passerella iliaca) Song Sparrow (Melospiza melodia) Lincoln’s Sparrow (Melospiza lincolnii) Swamp Sparrow (Melospiza georgiana) White-crowned Sparrow (Zonotrichia leucophrys) White-throated Sparrow (Zonotrichia albicollis) Harris’ Sparrow (Zonotrichia querula) Dark-eyed Junco (Junco hyemalis) White-winged race Slate-colored race Oregon race Gray-headed race McCown’s Longspur (Calcarius mccownii) Lapland Longspur (Calcarius lapponicus) Chestnut-collared Longspur (Calcarius ornatus) Snow Bunting (Plectrophenax nivalis) Bobolink (Dolichonyx oryzivorus) Western Meadowlark (Sturnella neglecta) Yellow-headed Blackbird (Xanthocephalus xanthocephalus) Red-winged Blackbird (Agelaius phoeniceus) Brewer’s Blackbird (Euphagus cyanocephalus) Common Grackle (Quiscalus quiscula) Brown-headed Cowbird (Molothrus ater) Orchard Oriole (Icterus spurius) Baltimore Oriole (Icterus galbula) Bullock’s Oriole (Icterus bullocki) Gray-crowned Rosy-finch (Leucosticte tephrocotis) Brown-capped Rosy-finch (Leucosticte australis) Black Rosy-finch (Leucosticte atrata) Pine Grosbeak (Pinicola enucleator) Cassin’s Finch (Carpodacus cassinii) House Finch (Carpodacus mexicanus) Red Crossbill (Loxia curvirostra) White-winged Crossbill (Loxia leucoptera)

M M S S S S S S S W M M S M S M M S S M S M Y M M W M W

R R C C U R R U C C A C C C C R R A C R C R C C R A U R

W W W W S W S W S Y S Y S S S S S S W W W W W Y W W

R U C U C C U U R C C C C C C C R C U R R R R C U R

208

Ecology of the Shortgrass Steppe

Common Redpoll (Carduelis flammea) Pine Siskin (Carduelis pinus) American Goldfinch (Carduelis tristis) Evening Grosbeak (Coccothraustes vespertinus) House Sparrow (Passer domesticus)

W Y Y W Y

U U C U A

References Andrén, H., and P. Angelstam. 1988. Elevated predation rates as an edge effect in habitat islands: Experimental evidence. Ecology 69:544–547. Andrén, H., P. Angelstam, E. Lindstrom, and P. Widen. 1985. Differences in predation pressure in relation to habitat fragmentation: an experiment. Oikos 45:273–277. Andrews, R., and R. Righter. 1992. Colorado birds. Denver Museum of Natural History, Denver, Colo. Askins, R. A. 1993. Population trends in grassland, shrubland and forest birds in eastern North America. Current Ornithology 11:1–34. Askins, R. A. 2000. Restoring North American birds: Lessons from landscape ecology. Yale University Press, New Haven, Conn. Baldwin, P. H. 1970. Avian food studies at the Pawnee site, pp. 84–85. In: R. T. Coupland and G. M. Van Dyne (eds.), Grassland ecosystems: Reviews of research. Range Science Department science series no. 7. Colorado State University, Fort Collins, Colo. Baldwin, P. H. 1971. Diet of the killdeer at the Pawnee National Grassland and a comparison with the Mountain Plover, 1970–1971. U.S. IBP Grassland Biome technical report no. 135. Colorado State University, Fort Collins, Colo. Baldwin, P. H., P. D. Creighton, and D. S. Kisiel. 1971. Diet of the mourning dove at the Pawnee National Grassland, 1970–1971. U.S. IBP Grassland Biome technical report no. 136. Colorado State University, Fort Collins, Colo. Basili, G. D., and S. A. Temple. 1999. Demographic characteristics of Dickcissels in winter. Studies in Avian Biology 19:281–288. Berthelsen, P. S., and L. M. Smith. 1995. Nongame bird nesting on CRP lands in the southern High Plains. Journal of Soil and Water Conservation 44:504–507. Best, L. B., H. Campa, K. E. Kemp, R. J. Robel, M. R. Ryan, J. A. Savidge, H. P. Weeks, and S. R. Winterstein. 1998. Avian abundance in CRP and crop fields during winter in the Midwest. American Midland Naturalist 139:311–324. Bock, C. E., V. A. Saab, T .D. Rich, and D. S. Dobkin. 1993. Effects of livestock grazing on neotropical migratory landbirds in western North America, pp. 296–309. In: D. M. Finch and P. W. Stangel (eds.), Status and management of neotropical migratory birds. USDA Forest Service general technical report RM-229. Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colo. Boren, J. C., D. M. Engle, and R. E. Masters. 1997. Vegetation cover type and avian species changes on landscapes within a wildland–urban interface. Ecological Modelling 103:251–266. Brennen, L. A., and W. P. Kulvesky. 2005. North American grassland birds: An unfolding conservation crisis? Journal of Wildlife Management 69:1–13. Cody, M. L. 1966. The consistency of intra- and inter-continental grassland bird species counts. American Naturalist 100:371–376. Cody, M. L. 1968. On the methods of resource division in grassland bird communities. American Naturalist 102:107–147. Cody, M. L. 1985. Habitat selection in grassland and open-country birds, pp. 191–226. In: M. L. Cody (ed.), Habitat selection in birds. Academic Press, San Diego, Calif.

Birds of the Shortgrass Steppe 209 Collins, S. L. 2001. Long-term research and the dynamics of bird populations and communities. Auk 118:583–588. Coppedge, B. R., D. M. Engle, R. E. Masters, and M. S. Gregory. 2001. Avian response to landscape change in fragmented southern Great Plains grasslands. Ecological Applications 11:47–59. Creighton, P. D., and P. H. Baldwin. 1974. Habitat exploitation by an avian ground-foraging guild. U.S. IBP Grassland Biome technical report no. 263. Colorado State University, Fort Collins, Colo. DeAngelis, D. L., and J. C. Waterhouse. 1987. Equilibrium and nonequilibrium concepts in ecological models. Ecological Monographs 57:1–21. Desmond, M. J., and J. A. Savidge. 1999. Satellite burrow use by Burrowing Owl chicks and its influence on nest fate. Studies in Avian Biology 19:128–130. Fewster, R. M., S. T. Buckland, G. M. Siriwardena, S. R. Baillie, and J. D. Wilson. 2000. Analysis of population trends for farmland birds using generalized additive models. Ecology 81:1970–1984. Fleischner, T. L. 1994. Ecological costs of livestock grazing in western North America. Conservation Biology 8:629–644. Fowler, A. C., R. L. Knight, T. L. George, and L. C. McEwen. 1991. Effects of avian predation on grasshopper populations in North Dakota grasslands. Ecology 72:1775–1781. Fretwell, S. D. 1972. Populations in a seasonal environment. Princeton University Press, Princeton, N.J. Gates, J. E., and L. W. Geysel. 1978. Avian nest dispersion and fledgling success in field– forest ecotones. Ecology 59:871–883. George, T. L., A. C. Fowler, R. L. Knight, and L. C. McEwen. 1992. Impacts of a severe drought on grassland birds in western North Dakota. Ecological Applications 2:275–284. Giezentanner, J. B. 1970. Avian distribution and population fluctuations on the shortgrass prairie of north–central Colorado. U.S. IBP Grassland Biome technical report no. 62. Colorado State University, Fort Collins, Colo. Giezentanner, J. B., and R. A. Ryder. 1969. Avian distribution and population fluctuations, Pawnee site. U.S. IBP Grassland Biome technical report no. 28. Colorado State University, Fort Collins, Colo. Golley, F. B. 1993. A history of the ecosystem concept in ecology: More than the sum of the parts. Yale University Press, New Haven, Conn. Graul, W. D. 1972. Breeding biology of the Mountain Plover (Charadrius montanus), 1969–1972. U.S. IBP Grassland Biome technical report no. 199. Colorado State University, Fort Collins, Colo. Greer, R. D. 1988. Effects of habitat structure and productivity on grassland birds. PhD diss., University of Wyoming, Laramie, Wyo. Hartnett, D. C., A. A. Steuter, and K. R. Hickman. 1997. Comparative ecology of native and introduced ungulates, pp. 72–101. In: F. L. Knopf and F. B. Samson (eds.), Ecology and conservation of Great Plains vertebrates. Springer, New York. Helzer, C. J. 1996. The effects of wet meadow fragmentation on grassland birds. Masters thesis, University of Nebraska, Lincoln, Neb. Herkert, J. R. 1994. The effects of habitat fragmentation on midwestern grassland bird communities. Ecological Applications 4:461–471. Herkert, J. R. 1995. An analysis of midwestern breeding bird population trends: 1966–1993. American Midland Naturalist 134:41–50. Hoekstra, J. M., T. M. Boucher, T. H. Ricketts, and C. Roberts. 2005. Confronting a biome crisis: Global disparities of habitat loss and protection. Ecology Letters 8:23–29.

210

Ecology of the Shortgrass Steppe

Howard, M. N., S. K. Skagen, and P. L. Kennedy. 2001. Does habitat fragmentation influence nest predation in the shortgrass prairie? Condor 103:530–536. Hughes, A. J. 1993. Breeding density and habitat preferences of the Burrowing Owl in northeastern Colorado. Masters thesis, Colorado State University, Fort Collins, Colo. James, F. C., C. E. McCulloch, and D. A. Wiedenfeld. 1996. New approaches to the analysis of population trends in land birds. Ecology 77:13–27. Joern, A., and K. H. Keeler. 1995. Getting the lay of the land: Introducing North American native grasslands, pp. 11–24. In: A. Joern and K. H. Keeler (eds.), The changing prairie: North American grasslands. Oxford University Press, New York. Johnson, D. H., and M. D. Schwartz. 1993. The Conservation Reserve Program and grassland birds. Conservation Biology 7:934–937. Knopf, F. L. 1994. Avian assemblages on altered grasslands. Studies in Avian Biology 15:232–246. Knopf, F. L. 1996. Prairie legacies: Birds, pp. 135–148. In: F. B. Samson and F. L. Knopf (eds.), Prairie conservation: Preserving North America’s most endangered ecosystem. Island Press, Covelo, Calif. Knopf, F. L., and B. J. Miller. 1994. Charadrius montanus: Montane, grassland, or bareground plover? Auk 111:505–506. Knopf, F. L., and J. R. Rupert. 1995. Habits and habitats of Mountain Plovers in winter. Condor 97:743–751. Knopf, F. L., and J. R. Rupert. 1996. Reproduction and movements of Mountain Plovers breeding in Colorado. Wilson Bulletin 108:28–35. Knopf, F. L., and J. R. Rupert. 1999. Use of cultivated fields by breeding Mountain Plovers in Colorado. Studies in Avian Biology 19:81–86. Knopf, F. L., and F. B. Samson. 1997. Conservation of grassland vertebrates, pp. 273–289. In: F. L. Knopf and F. B. Samson (eds.), Ecology and conservation of Great Plains vertebrates. Springer, New York. Küchler, A. W. 1964. Potential natural vegetation of the coterminous United States. Special publication 36. American Geographical Society, New York. Laubhan, M. K., and L. H. Fredrickson. 1997. Wetlands of the Great Plains: Habitat characteristics and vertebrate aggregations, pp. 20–48. In: F. L. Knopf and F. B. Samson (eds.), Ecology and conservation of Great Plains vertebrates. Springer, New York. Lauenroth, W. K., and I. C. Burke. 1995. Great Plains, climate variability, pp. 237–249. In: W. A. Nierenberg (ed.), Encyclopedia of environmental biology. Vol. 2. Academic Press, New York. Lauenroth, W. K., and D. G. Milchunas. 1991. Shortgrass steppe, pp. 183–226. In: R. T. Coupland (ed.), Ecosystems of the world 8A: Natural grasslands. Elsevier, Amsterdam. Lauenroth, W. K., and O. E. Sala. 1992. Long term forage production of North American shortgrass steppe. Ecological Applications 2:397–403. Link, W. A., and J. R. Sauer. 1997. New approaches to the analysis of population trends in land birds: Comment. Ecology 78:2632–2634. Link, W. A., and J. R. Sauer. 1998. Estimating population change from count data: Applications to the North American Breeding Bird Survey. Ecological Applications 8:258–268. Marti, C. D. 1969. Some comparisons of the feeding ecology of four owls in north–central Colorado. Southwestern Naturalist 14:163–170. McCoy, T. D., M. R. Ryan, E. W. Kurzejeski, and L. W. Burger. 1999. Conservation Reserve Program: Source or sink habitat for grassland birds in Missouri? Journal of Wildlife Management 63:530–538.

Birds of the Shortgrass Steppe 211 McCullough, D. R. 1970. Secondary production of birds and mammals, pp. 107–130. In: D. Reichle (ed.), Analysis of temperate forest ecosystems. Springer-Verlag, New York. McIntyre, N. E., and T. R. Thompson. 2003. A comparison of Conservation Reserve Program habitat plantings with respect to arthropod prey for grassland birds. American Midland Naturalist 150:291–301. Miller, B. J., and F. L. Knopf. 1993. Growth and survival of Mountain Plovers. Journal of Field Ornithology 64:500–506 and 65:193. O’Connor, R .J., M. T. Jones, R. B. Boone, and T. B. Lauber. 1999. Linking continental climate, land use, and land patterns with grassland bird distribution across the conterminous United States. Studies in Avian Biology 19:45–59. Olendorff, R. R. 1972. The large birds of prey of the Pawnee National Grassland: Nesting habits and productivity, 1969–1971. U.S. IBP Grassland Biome technical report no. 151. Colorado State University, Fort Collins, Colo. Olendorff, R. R. 1973. The ecology of the nesting birds of prey of northeastern Colorado. U.S. IBP Grassland Biome technical report no. 211. Colorado State University, Fort Collins, Colo. Parton, W. J., D. S. Ojima, and D. S. Schimel. 1994. Environmental change in grasslands: Assessment using models. Climatic Change 28:111–141. Peterjohn, B. G., and J. R. Sauer. 1999. Population status of North American grassland birds from the North American Breeding Bird Survey, 1966–1996. Studies in Avian Biology 19:27–44. Peterjohn, B. G., J. R. Sauer, and C. S. Robbins. 1995. Population trends from the North American Breeding Bird Survey, pp. 3–39. In: T. E. Martin and D. M. Finch (eds.), Ecology and management of neotropical migratory birds: A synthesis and review of critical issues. Oxford University Press, New York. Pickett, S. T. A., J. Kolasa, and C. G. Jones. 1994. Ecological understanding: The nature of theory and the theory of nature. Academic Press, Orlando, Fla. Pleszczynska, W. K. 1977. Polygyny in the Lark Bunting. PhD diss., University of Toronto, Toronto, Ont. Plumpton, D. L. 1992. Aspects of nest site selection and habitat use by Burrowing Owls at the Rocky Mountain Arsenal, Colorado. Masters thesis, Texas Tech University, Lubbock, Texas. Porter, D. K. 1974. Accuracy in censusing breeding passerines on the shortgrass prairie. U.S. IBP Grassland Biome technical report no. 254. Colorado State University, Fort Collins, Colo. Reynolds, R. E., T. L. Shaffer, J. R. Sauer, and B. G. Peterjohn. 1994. Conservation Reserve Program: Benefit for grassland birds in the northern Plains. Transactions of the North American Wildlife Natural Resources Conference 59:328–336. Rosenzweig, M. L. 1968. Net primary productivity of terrestrial communities: Predictions from climatological data. American Naturalist 102:67–74. Rotenberry, J. T. 1985. The role of habitat in avian community composition: Physiognomy or floristics? Oecologia 67:213–217. Rotenberry, J. T., and J. A. Wiens. 1980a. Habitat structure, patchiness, and avian communities in North American steppe vegetation: A multivariate approach. Ecology 61:1228–1250. Rotenberry, J. T., and J. A. Wiens. 1980b. Temporal variation in habitat structure and shrubsteppe bird dynamics. Oecologia 47:1–9. Ryder, R. A. 1970. Avian populations at the Pawnee site, pp. 84–85. In: R. T. Coupland and G. M. Van Dyne (eds.), Grassland ecosystems: Review of research. Range Science Department science series no. 7. Colorado State University, Fort Collins, Colo.

212

Ecology of the Shortgrass Steppe

Ryder, R. A. 1972. Avian population studies on the Pawnee site, 1968–1971. U.S. IBP Grassland Biome technical report no. 171. Colorado State University, Fort Collins, Colo. Ryder, R. A. 1980. Effects of grazing on bird habitats, pp. 51–66. In: R. M. Degraff and N. G. Tilghman (eds.), Management of western forests and grasslands for nongame birds. USDA Forest Service general technical report INT-86. Intermountain Forest and Range Experiment Station, Ogden, Utah. Ryder, R. A., and D. A. Cobb. 1969. Birds of the Pawnee National Grassland in northern Colorado. Journal of the Colorado–Wyoming Academy of Science 6:5. (Abst.). Samson, F. B., and F. L. Knopf. 1994. Prairie conservation in North America. BioScience 44:418–421. Sauer, J. R., J. E. Hines, and J. Fallon. 2005. The North American Breeding Bird Survey, results and analysis 1966–2005, version 6.2.2006. USGS Patuxent Wildlife Research Center, Laurel, Md. Also available at http://www.mbr-pwrc.usgs.gov/ bbs/bbs.htm. Sherry, T. W., and R. T. Holmes. 1996. Winter habitat quality, population limitation, and conservation of neotropical–neoarctic migrant birds. Ecology 77:36–48. Skagen, S. K. 1997. Stopover ecology of transitory populations: The case of migrant shorebirds, pp. 244–269. In: F. L. Knopf and F. B. Samson (eds.), Ecology and conservation of Great Plains vertebrates. Springer, New York. Skagen, S. K., A. A. Yackel Adams, and R. D. Adams. 2005. Nest survival relative to patch size in a highly fragmented shortgrass prairie landscape. Wilson Bulletin 117:23–34. Strong, M. A. 1971. Avian productivity on the shortgrass prairie of north–central Colorado. Masters thesis, Colorado State University, Fort Collins, Colo. Temple, S. A., B. M. Fevold, L. K. Paine, D. J. Undersander, and D. W. Sample. 1999. Nesting birds and grazing cattle: Accommodating both on midwestern pastures. Studies in Avian Biology 19:196–202. Thomas, L. 1996. Monitoring long-term population change: Why are there so many analysis methods? Ecology 77:49–58. Thomas, L., and K. Martin. 1996. The importance of analysis method for Breeding Bird Survey population trend estimates. Conservation Biology 10:479–490. Thompson, T. R. 2003. The effectiveness of the Conservation Reserve Program’s native seeding requirement in providing breeding and wintering habitat for grassland birds in the southern High Plains of Texas. Masters thesis, Texas Tech University, Lubbock, Texas. Vickery, P. D., and J. R. Herkert. 2001. Recent advances in grassland bird research: Where do we go from here? Auk 188:11–15. Vickery, P. D., P. L. Tubaro, J. M. Cardoso da Silva, B. G. Peterjohn, J. R. Herkert, and R. B. Cavalcanti. 1999. Conservation of grassland birds in the western hemisphere. Studies in Avian Biology 19:2–26. Walk, J. W., and R. E. Warner. 1999. Effects of habitat area on the occurrence of grassland birds in Illinois. American Midland Naturalist 141:339–344. Wiens, J. A. 1969. An approach to the study of ecological relationships among grassland birds. Ornithological Monographs 8:1–93. Wiens, J. A. 1970a. Avian populations and patterns of habitat occupancy at the Pawnee site, 1968–1969. U.S. IBP Grassland Biome technical report no. 63. Colorado State University, Fort Collins, Colo. Wiens, J. A. 1970b. Habitat heterogeneity and avian consumer populations in grasslands, pp. 77–83. In: R. T. Coupland and G. M. Van Dyne (eds.), Grassland ecosystems:

Birds of the Shortgrass Steppe 213 Reviews of research. Range Science Department science series no. 7. Colorado State University, Fort Collins, Colo. Wiens, J. A. 1971. Avian ecology and distribution in the Comprehensive Network, 1970. U.S. IBP Grassland Biome technical report no. 77. Colorado State University, Fort Collins, Colo. Wiens, J. A. 1973. Pattern and process in grassland bird communities. Ecological Monographs 43:237–270. Wiens, J. A. 1974a. Climatic instability and the “ecological saturation” of bird communities in North American grasslands. Condor 76:385–400. Wiens, J. A. 1974b. Habitat heterogeneity and avian community structure in North American grasslands. American Midland Naturalist 91:195–213. Wiens, J. A. 1975. Rangeland avifaunas: Their composition, energetics, and role in the ecosystem, pp. 146–182. In: D. R. Smith (tech. coord.), Proceedings of the Symposium on Management of Forest and Range Habitats for Nongame Birds. USDA Forest Service general technical report WO-1. Washington, D.C. Wiens, J. A. 1977. On competition and variable environments. American Scientist 65:590–597. Wiens, J. A. 1978. Nongame bird communities in northwestern coniferous forests, pp. 19–20. In: R. M. DeGraaf (ed.), Proceedings of the Workshop on Nongame Bird Habitat Management in the Coniferous Forests of the Western United States. USDA Forest Service general technical report PNWS-64. Portland, Ore. Wiens, J. A. 1981. Single-sample surveys of communities: Are the revealed patterns real? American Naturalist 117:90–98. Wiens, J. A. 1984. The place of long-term studies in ornithology. Auk 101:202–203. Wiens, J. A., and M. I. Dyer. 1975. Rangeland avifaunas: Their composition, energetics and role in the ecosystem, pp. 146–182. In: D. R. Smith (tech. coord.), Proceedings of the Symposium on Management of Forest and Range Habitats for Nongame Birds. USDA Forest Service general technical report WO-1. Washington, D.C. Wiens, J. A., and G. S. Innis. 1974. Estimation of energy flow in bird communities: Population bioenergetics model. Ecology 55:730–746. Wiens, J. A., and J. T. Rotenberry. 1979. Diet niche relationships among North American grassland and shrubsteppe birds. Oecologia 42:253–292. Wiens, J. A., and J. T. Rotenberry. 1981. Habitat associations and community structure of birds in shrubsteppe environments. Ecological Monographs 51:21–41. Wiens, J. A., J. T. Rotenberry, and J. F. Ward. 1972. Avian populations at IBP Grassland Biome sites: 1971. U.S. IBP Grassland Biome technical report no. 205. Colorado State University, Fort Collins, Colo. Wiens, J. A., J. T. Rotenberry, and J. F. Ward. 1974a. Bird populations at ALE, Pantex, Osage, and Cottonwood, 1972. U.S. IBP Grassland Biome technical report no. 267. Colorado State University, Fort Collins, Colo. Wiens, J. A., J. F. Ward, and J. T. Rotenberry. 1974b. Dietary composition and relationships among breeding bird populations as US/IBP Grassland Biome sites, 1970. U.S. IBP Grassland Biome technical report no. 262. Colorado State University, Fort Collins, Colo. Wilcove, D. S. 1985. Nest predation in forest tracts and the decline of migratory songbirds. Ecology 66:1211–1214. Winter, M., and J. Faaborg. 1999. Patterns of area sensitivity in grassland-nesting birds. Conservation Biology 13:1424–1436. With, K. A. 1994. The hazards of nesting near shrubs for a grassland bird, the McCown’s Longspur. Condor 96:1009–1019.

214

Ecology of the Shortgrass Steppe

With, K. A., and D. R. Webb. 1993. Microclimate of ground nests: The relative importance of radiative cover and wind breaks for three grassland species. Condor 95:401–413. Yackel Adams, A. A., S. K. Skagen, and R. D. Adams. 2001. Movements and survival of Lark Bunting fledglings. Condor 103:643–647. Zimmerman, G., P. Stapp, and B. Van Horne. 1996. Seasonal variation in the diet of Great Horned Owls (Bubo virginianus) on shortgrass prairie. American Midland Naturalist 136:149–156.

10 Insect Populations, Community Interactions, and Ecosystem Processes in the Shortgrass Steppe Thomas O. Crist

I

nsects are diverse, abundant, and have numerous roles in rangeland ecosystems. More than 1600 species representing 238 families of insects have been recorded in the shortgrass steppe of northeastern Colorado (Kumar et al., 1976). Of this large assemblage, a much smaller subset—perhaps fewer than 50 species—is highly abundant with a large influence on community and ecosystem processes (Lauenroth and Milchunas, 1992). Even within abundant insect groups, such as grasshoppers, some species have far greater effects than others as herbivores (Capinera, 1987). In this chapter I consider a small number of insect groups that have various influences in shortgrass steppe ecosystems (Table 10.1). I focus on three insect taxa—grasshoppers, beetles, and ants—that are widespread, abundant, and ecologically important in semiarid environments. I also draw attention to neglected groups, such as termites and spiders, for their potentially important roles in the shortgrass steppe. My primary objective is to emphasize the linkages among insect populations, community interactions, and ecosystem function. From this approach stems several related issues: how population distributions affect community interactions, how population abundance affects the processing and redistribution of energy and nutrients in ecosystems, and how abundance and species diversity are important to the functional roles of species in ecosystems. I skirt issues of population regulation in insects, which are reviewed elsewhere (Cappuccino and Price, 1995), and instead consider how temporal and spatial patterns in insect populations relate to community and ecosystem processes. Understanding relationships among populations, communities, and ecosystems requires approaches that link patterns and processes across scales. Much of what

215

216

Ecology of the Shortgrass Steppe

Table 10.1 Some Abundant Insects and Other Arthropods with Important Roles in the Shortgrass Steppe Order and Family

Genera

Feeding Mode

Potential Roles in Ecosystem

Araneae Gnaphosidae

Gnaphosa

Predator

Lycosidae

Schizocosa

Predator

Effects on prey, role in food webs Effects on prey, role in food webs

Harpalus Pasimachus

Omnivore Predator

Scarabaeidae

Phyllophaga

Herbivore

Tenebrionidae

Eleodes

Detritivore

Efferia

Predator

Effects on prey, role in food webs

Formica Pogonomyrmex

Predator Granivore

Role in food webs, soils Disturbance, seed pool, soils

Isoptera Rhinotermitidae

Reticulitermes

Decomposer

Nutrient cycling, soils

Orthoptera Acrididae

Opeia

Herbivore

Effects on primary production, litter production, food webs

Melanoplus

Herbivore

Coleoptera Carabidae

Diptera Asilidae Hymenoptera Formicidae

Plants/prey, role in food webs Effects on prey, role in food webs Disturbance, primary production Litter processing and redistribution

is known about the roles of insects in the shortgrass steppe is based on studies conducted at relatively fine scales. To link insect population studies to community and ecosystem processes, however, I suggest that insect populations should also be studied across broader scales that encompass topographic variation. The rolling topography in the shortgrass steppe produces a gradient in soil texture, water availability, and nutrient retention from uplands to lowlands (Clark and Woodmansee, 1992; Schimel et al., 1985). Plant community structure also varies with topography in spatially repeating patterns across the landscape (Milchunas et al., 1989). We know considerably less about how insect distribution and abundance changes along topographic gradients, and how these changes might influence community and ecosystem processes. I discuss ways in which insect distributions might be examined across a range of scales in the shortgrass steppe. Much of this review is based on studies conducted at the CPER in northeastern Colorado. I also draw upon findings from other shortgrass steppe, shrub steppe or mixed-grass ecosystems, to place the CPER studies in a broader regional context.

Insect Populations, Community Interactions, and Ecosystem Processes 217

Population Dynamics and Distribution Casual observation of insects inhabiting the shortgrass steppe suggests that ground-dwelling beetles (Carabidae, Scarabaeidae, Tenebrionidae; Fig. 10.1), grasshoppers (Acrididae), and ants (Formicidae) are abundant and widely distributed insect groups. In species richness, biomass, and abundance, these taxa may represent a significant fraction of the insect fauna in the shortgrass steppe (Kumar et al., 1976; Lauenroth and Milchunas, 1992; Lavigne and Kumar, 1974). Two taxa, grasshoppers and harvester ants (Pogonomyrmex spp.; Fig. 10.2), have long been considered rangeland pests in shortgrass ecosystems because of their abundance and impacts on vegetation and soils (Capinera, 1987; Rogers, 1987). Spiders (Araneae) are also abundant in the shortgrass steppe, but few studies have been conducted to examine their roles as predators in food webs (Lavigne and Kumar, 1974; Weeks and Holtzer, 2000). Similarly, the arid-land subterranean termite (Reticulitermes tibialis) is surprisingly common in lowland areas, where they may have an important role as decomposers of woody plant litter (Crist, 1998). These arthropod groups vary substantially in patterns of distribution and abundance. They also have different feeding modes and functional roles in the shortgrass steppe (Table 10.1). The combined effects of insects on ecosystem functioning may be considerable. The major challenges in assessing these roles are difficulties in extrapolating variable patterns of abundance across time and space, and inferring processes from a complex set of patterns.

Figure 10.1 Several species of darkling beetles (Coleoptera: Tenebrionidae) occur in the shortgrass steppe. Pictured here is the large-bodied Eleodes hispilabris. (Photo by T. O. Crist.)

218

Ecology of the Shortgrass Steppe

Figure 10.2 Nest mound and clearing of the Western harvester ant (Pogonomyrmex occidentalis). Ant colony densities may exceed 30/ha in the shortgrass steppe. (Photo by T. O. Crist.)

Temporal Dynamics Grasshopper assemblages from various shortgrass sites typically have 25 to 40 species (Onsager, 1987; Przybyszewski and Capinera, 1990; Welch et al., 1991), with three to five species representing most of the variation in overall abundance among years (Capinera and Thompson, 1987; Onsager, 1987; Przybyszewski and Capinera, 1990; Van Horn, 1970; Welch et al., 1991). Many of the dominant grasshopper species in shortgrass are in the subfamily Gomphocerinae, which tend to be more abundant in warmer, drier regions dominated by C4 grasses. Species of Melanoplinae are more abundant in mixed-grass prairie where C3 grasses predominate (Belovsky and Joern, 1995; Capinera, 1987). Oedipodinae are locally abundant in more open areas with bare ground (Przybyszewski and Capinera, 1990; Uvarov, 1977). At the CPER, for example, Welch et al. (1991) found that 10 of 34 species comprised 95% of the total grasshopper abundance, four of which were gomphocerines; five, oedipodines; and one, melanopline. Thus, grasshopper species that have large population fluctuations or outbreaks vary among regions according to subfamilies. A long history of grasshopper outbreaks is documented in western rangelands (Pfadt and Hardy, 1987). During outbreaks, population densities of some species often exceed 20 ∙ m–2 and can remove 20% to 35% of the standing crop of vegetation (Hewitt and Onsager, 1983; Mitchell and Pfadt, 1974; Onsager, 1987). Severe

Insect Populations, Community Interactions, and Ecosystem Processes 219

outbreaks may last from 2 to 6 years, during which much of the green plant biomass may be consumed by grasshoppers (Pfadt and Hardy, 1987). The timing and spatial patterning of outbreaks are variable and some areas have a tendency to outbreak more than others. Lockwood and Lockwood (1991), for instance, report that most outbreaks in Wyoming occur primarily in the northern mixed-grass prairie. Studies at the CPER show that grasshopper densities are usually less than 10 ∙ m–2 (a value often considered outbreak density) (Lockwood and Lockwood, 1991; Torell and Huddleston, 1987) but often exceed 3.6 ∙ m–2 (the threshold infestation level) (Capinera and Horton, 1989). From 1980 to 1985, Capinera and Thompson (1987) found that mean densities of grasshoppers in eight pastures changed 10-fold from 0.48 to 4.83 ∙ m–2. Densities within some pastures were more variable among years, ranging from 0.14 to 15.4 ∙ m–2. Two species—Opeia obscura (Gomphocerinae) and Melanoplus gladstoni (Melanoplinae)—were primarily responsible for changes in overall grasshopper densities, and these species had their peak abundances during consecutive years (Capinera and Thompson, 1987). Darkling beetles (Tenebrionidae) and ground beetles (Carabidae) also exhibit considerable yearly variation in abundance in the shortgrass steppe. Crist and Wiens (unpublished data) recorded beetle abundances from 1990 to 1994 along a 900-m pitfall transect (trap spacing, 10 m) spanning a topographic gradient. As a group, tenebrionids had low abundances in 1990 and 1991, but showed an order of magnitude increase in 1992 and maintained high levels in 1993 and 1994 (Fig. 10.3A). Changes in overall tenebrionid abundances were primarily the result of Eleodes extricata and E. obsoleta (Fig. 10.3A). These two species, along with E. hispilabris, comprised most of the total abundance of the 15 tenebrionid species recorded during the study. Between 1990 and 1992, E. obsoleta was most abundant, and seasonal increases in E. obsoleta (a late-season species) closely tracked overall tenebrionid abundance. In 1993 and 1994, however, E. extricata became more common, and seasonal decreases in tenebrionid abundances were primarily the result of E. extricata (an early-season species). This 5-year trend in tenebrionid abundance can be extended with McIntyre’s (2000) 4-year study, which showed that densities of E. extricata and E. hispilabris were high in 1994 (consistent with Fig. 10.3A) and declined steadily until 1997. Decreases in beetle abundances corresponded to a shift from warmer, drier conditions in 1994 and 1995 to cooler, wetter weather in 1996 and 1997. Beetle densities were influenced by temperature and precipitation within a year as well as from the previous year (McIntyre, 2000). Carabid beetles showed similar yearly patterns of abundance to tenebrionids along the 900-m transect during 1990 to 1994. The overall carabid beetle abundance also increased during 1992 and 1993, but decreased in 1994 (Fig. 10.3B). Carabids consistently increased in abundance from early to late summer. Grounddwelling spiders showed abundances in pitfall traps that were comparable with both groups of beetles (Fig. 10.3B). More than 80% of the spiders captured in pitfall traps are wolf spiders (Lycosidae: Schizocosa spp.) or hunting spiders (Gnaphosidae) (Weeks, 1996; Weeks and Holtzer, 2000). As a group, grounddwelling spiders appeared to have lower interannual variability in abundance than tenebrionid or carabid beetles.

220

Ecology of the Shortgrass Steppe 150

a)

125

E. hispilabris E. extricata

100

E. obsoleta All Tenebrionidae

Mean Abundance (captures day-1)

75 50 25 0 Jul Aug

Jun Jul Aug

Jul Aug

Jun Jul Aug

Jun Jul Aug

Jul Aug

Jun Jul Aug

Jun Jul Aug

1993

1994

150

b) Carabidae Araneae

125 100 75 50 25 0 Jul Aug

1990

Jun Jul Aug

1991

1992

Figure 10.3 Patterns of insect abundance from 1990 to 1994 along a 900-m transect that spans a topographic gradient. Data are expressed as total number of individuals captured per day (based on 3–6 days of trapping) for each sample month. (A) Overall abundance of darkling beetles (Tenebrionidae) and dynamics of three Eleodes species. (B) Overall abundance of ground beetles (Carabidae) and ground-dwelling spiders (Araneae).

Densities of harvester ant colonies are reported in several studies (Crist and Wiens, 1996; Rogers and Lavigne, 1974), but few data are available on population dynamics. Porter and Jorgensen (1988) found colony densities of P. owyheei remained fairly constant at 40 colonies ∙ ha–1 from 1977 to 1986, and 80% of the nests sampled in 1977 were still active in 1986. Keeler (1993) studied 15-year colony dynamics of P. occidentalis in western Nebraska. Between 1977 and 1991, colony density varied from 56 to 80 colonies ∙ ha–1. The recruitment of new colonies was 0 to 8 colonies ∙ year –1 and the annual mortality was 0 to 5 colonies ∙ year –1. Despite yearly turnover, the population was highly stable from 1982 to 1991, fluctuating narrowly from 74 to 80 colonies ∙ ha–1. From these data there emerge clear differences among insect taxa in their annual variation in abundance. Grasshoppers are most variable, and their dynamics are

Insect Populations, Community Interactions, and Ecosystem Processes 221

more directly tied to annual variation in weather and aboveground primary production (e.g., Belovsky and Slade, 1995) than those of other groups. Tenebrionids are also highly variable in population size, but because of their largely detritivorous food habits, they appear to be less influenced by annual variation in primary production than grasshoppers. The relatively slow development and long life span of tenebrionids (2–3 years) (Allsopp, 1980), however, may delay their responses to food availability or weather. The corresponding trends in yearly abundances of tenebrionids and carabids suggest that similar factors such as annual variation in temperature and precipitation affect beetle abundances (McIntyre, 2000). The most stable population dynamics are shown by harvester ant colonies, which are long-lived (10–35 years) after they become established (Keeler, 1993; Porter and Jorgensen, 1988). Among the insect groups, therefore, greater yearly variability in abundance appears to be associated with shorter life span and generation time (grasshoppers), whereas more stable dynamics are found in longer lived species (carabids, tenebrionids, and especially ants). Spatial Distribution Microhabitat Patterns Patterns of insect movement and microhabitat use can provide a mechanistic basis for understanding population distribution (Crist and Wiens, 1995; Johnson et al., 1992a, b; Wiens et al., 1993b; With and Crist, 1995). Studies of insect movement and microhabitat use, however, are often limited by time-intensive and specialized methodology (Turchin, 1998; Turchin et al., 1991; Wiens et al., 1993a). Joern (1983b) measured the daily movements of four grasshopper species in the shortgrass steppe of west Texas. Grasshopper movement rates and net displacements were quite similar among species. The average distances moved by individuals were 4.9 to 8.1 m⋅d–1 and 40 to 70 m after 14 days. Species differed in their patterns of movement, depending on their vagility and microhabitat specificity (Joern, 1982, 1983b). With and Crist (1995) examined fine-scale movements of two grasshopper species—Psolessa delicatula (Gomphocerinae) and Xanthoppus corallipes (Oedopodinae)—at the CPER. Movement rates of each species were measured in vegetation patches that were classified as homogenous, moderately heterogeneous, or very heterogeneous, based on the degree to which Bouteloua gracilis was interspersed with other plant cover types such as cacti, shrubs, and forbs. Psolessa delicatula is small bodied, has limited vagility, and prefers homogenous patches of B. gracilis, whereas X. corallipes is larger, more vagile, and prefers more heterogeneous areas (With and Crist, 1995). A simulation model was then used to predict transition probabilities among patches using movement rates measured in the field. Model predictions were consistent with observed grasshopper distributions in the field. Psolessa delicatula showed a random distribution and was most abundant in homogenous patches, whereas X. corallipes was highly aggregated in heterogeneous patches of vegetation (With and Crist, 1995). Studies have also examined tenebrionid beetle movements in relation to habitat structure in the shortgrass steppe. Beetle movements were measured in areas that

222

Ecology of the Shortgrass Steppe

differed in vegetation structure, based on the proportions of grass, bare ground, cacti, and shrubs (Crist and Wiens, 1995; Crist et al., 1992). Rates of movement of Eleodes beetles differed among species but were generally greater in the more homogenous grass and bare ground than in heterogeneous cactus and shrub areas. Movement patterns among Eleodes species and habitats were nonetheless quite similar when displacements were appropriately rescaled to the overall length of the movement pathway (Crist et al., 1992). An important consequence of differential movement rates is that the population density of some Eleodes beetles may be three to four times greater in areas of cactus and shrub than in more homogeneous vegetation because of the greater residence times in cactus–shrub microhabitats (Crist and Wiens, 1995; McIntyre, 1997, 2000; Stapp, 1997). Broad-Scale Patterns Broad-scale patterns of insect distribution in the shortgrass steppe can be related to variation in soils and vegetation as they are influenced by topography and grazing. In two respects, the broad-scale spatial patterns are the most critical bits of information in linking insect populations to rangeland ecosystem processes. First, there is considerable broad-scale turnover of species composition and relative abundance among grassland insects and plants as a result of topography or land use (Tscharntke and Greiler, 1995). Second, many ecosystem processes, such as primary production and soil nutrient dynamics, are driven by topography or grazing (Lauenroth and Milchunas, 1992). For these reasons, a major gap in our understanding of the roles of insects in shortgrass steppe ecosystems stems from our lack of information on the broad-scale spatial distributions of insects. Grasshoppers are strongly influenced by broad-scale changes in vegetation (Kemp, 1992; Kemp et al., 1990; Miller and Onsager, 1991). Przybyszewski and Capinera (1990) found that several grasshopper species reached their highest abundances in different pastures, reflecting variation in vegetation among pastures (resulting from topography, soils, and/or grazing). In an earlier study, Capinera and Sechrist (1982) analyzed grasshopper abundance and plant biomass in six pastures with different grazing intensities. Overall grasshopper densities were 1.35 ∙ m–2 in high-biomass pastures (which were ungrazed or lightly grazed) and 0.75 ∙ m –2 in low-biomass pastures (moderately or heavily grazed). Population responses to grazing differed among subfamilies. As grazing intensity increased, gomphocerines decreased by half, melanoplines were similar in abundance, and oedopodines increased. Abundances of Gomphocerinae and Melanoplinae were positively correlated with the biomass of grasses, forbs, and shrubs, whereas Oedipodinae was negatively correlated with all three plant life-forms. Capinera and Sechrist (1982) noted that the observed differences in plant biomass resulting from grazing might be commonly encountered within smaller areas and can result in significant finescale changes in species abundances of grasshoppers—a point that deserves further study. Grasshopper responses to grazing can also differ over time. Welch et al. (1991) conducted grasshopper sampling in lightly and heavily grazed pastures and compared their results with those of Van Horn (1970). In the heavily grazed pasture, densities were similar between time periods: 0.86 ∙ m–2 and 0.78 ∙ m–2 in 1970 and 1989, respectively. In the lightly grazed pasture, however, densities were

Insect Populations, Community Interactions, and Ecosystem Processes 223

0.99 ∙ m–2 in 1970 and 1.44 ∙ m–2 in 1989. Heavily grazed pastures had significantly lower grasshopper densities in 1989 but not in 1970; grazing effects were therefore more apparent when viewed over longer time intervals (Welch et al., 1991). Broad-scale variation in rangeland is often related to topography, especially in areas such as the shortgrass steppe or mixed-grass prairie, where rolling topography gives rise to spatially repeating patterns of vegetation (Barnes and Harrison, 1982; Milchunas et al., 1989). Landscape patterns of vegetation and soils resulting from topographic variation are central to community and ecosystem studies. Soils generally change from coarse textured in uplands to fine textured in lowlands (Clark and Woodmansee, 1992). Corresponding increases in soil nutrient levels in lowlands (Schimel et al., 1985) produce increased plant biomass and changes in plant species composition (Milchunas et al., 1989). Insect species abundances should also change across a toposequence depending on species responses to changes in soils, vegetation, and/or slope exposure. A sampling of several insect species across a topographic gradient at the CPER (Fig. 10.4A) reveals distinct patterns of species distributions. The spatial

(B)

(A)

167

6

16

61

16 55

16

61

0.5 km

N

Figure 10.4 (A) Topography within a pasture at the CPER (3-m contour lines) where a 900-m pitfall transect (dotted line) was placed to conduct broad-scale measures of insect distribution and abundance. (B) Map of colonies of harvester ants (Pogonomyrmex occidentalis) obtained from digitizing aerial photographs (1:2000 scale) in the same pasture. The three belt transects sectioned into 1-ha segments illustrate how patterns of colony density across the toposequence can differ horizontally (see Fig. 10.6 for analysis).

224

Ecology of the Shortgrass Steppe

distribution of harvester ant colonies also varies along the same topographic gradient (Fig. 10.4B). Crist and Wiens (1996) conducted a more extensive mapping of harvester ant colonies (P. occidentalis) in five pastures at the CPER using low-level aerial photographs (1:2000 scale). Topography, soils, and grazing intensity influenced ant colony density and spacing patterns. Ants were most abundant on upland plains with coarse-textured soils and in pastures with light or moderate grazing intensity; colony density often exceeded 30 colonies ∙ ha–1 on the upland plain, but was generally less than 10 colonies ∙ ha–1 in lowland areas (Crist and Wiens, 1996). Across the same toposequence, E. extricata showed higher capture rates in pitfall traps on upper slopes, where a relatively greater proportion of bare ground interrupts the grass matrix; fewer pitfall captures were recorded in the intervening swale (Fig. 10.5A, B). Variation in pitfall captures was also clearly expressed at finer scales, especially in 1993 to 1994, when abundance was higher than in 1991. Although overall abundance of beetles varied considerably among years, the broad-scale spatial patterning was relatively consistent (Fig. 10.3B). Wood ants (Formica obscuripes) had a more patchy distribution (Fig. 10.5D). Captures of ants were greatest in the swale where shrubs are more abundant (Fig. 10.5C), partly because F. obscuripes uses woody debris for nest construction. In another study, Bestelmeyer and Wiens (2001) recorded a positive spatial association between F. obscuripes and saltbush (Atriplex canescens) habitats. The arid-land subterranean termite (R. tibialis) also had higher abundance in saltbush swales (Fig. 10.5E). A detailed geostatistical analysis showed that termite spatial distribution was strongly associated with saltbush (Crist, 1998). The peak in the probability of termite occurrence, however, was shifted slightly toward the south-facing slope relative to that of shrub proximity; therefore, soils or microclimate also likely affect termite distributions (Fig. 10.5C, E [see also Crist, 1998]). The spatial scales of variability in abundance of insect distributions can be compared by autocorrelation. Eleodes extricata exhibited a high degree of autocorrelation at 10-m and 80-m lags, indicating patchiness in pitfall samples within a topographic position (Fig. 10.6A). A stronger spatial patterning occurred in years with greater abundance (1993 and 1994 compared with 1991). Changes in beetle abundance across slope positions are reflected in the switch from positive to negative autocorrelation (ca., 200 m). Formica obscuripes also had a high correlation of ant numbers in adjacent traps with another distinct peak at 120 m in 1993 to 1994 and at 70 m in 1991 (Fig. 10.6B). There is a greater similarity in spatial pattern among years for F. obscuripes than for E. extricata. This might be expected because F. obscuripes is a social insect with long-lived colonies. Cross-correlation can be used to measure shifts in the spatial pattern of abundance among years. The degree of correlation in E. extricata abundance between 1991 and 1993 was higher among traps spaced 50 to 100 m apart (lags, –50 to –100 m) than in the same traps (zero lag) (Fig. 10.7A), indicating shifts in spatial distribution between years. A similar pattern occurred between 1991 and 1994. The cross-correlation between 1993 and 1994, however, was greatest near zero, indicating a similar spatial pattern of captures between these 2 years (Fig. 10.7A). Other studies have shown that landscape patterns of tenebrionid beetles are related to topographic variation in soils and vegetation (Bossenbroek et al., 2004; Crist

Proportion

1.0

(A) Vegetation Cover

Bare

Grass

Cactus

Shrub

0.8 0.6 0.4 0.2

# captured day-1

0.0

(B) Abundance of Eleodes extricata

2.0

91 93 94

1.5 1.0 0.5 0.0

(C) Distance to nearest shrub

All shrubs

Atriplex canescens

Distance (m)

10.0 8.0 6.0 4.0 2.0 0.0

# captured day-1

10.0

# of termites

20.0 (D) Abundance of Formica obscuripes

1800 1500 1200 900 600 300 0

15.0

91 93 94

5.0 0.0

(e) Abundance of Reticulitermes tibalis

0

100

200

300

400

500

600

700

800

900

Distance along transect (m) Figure 10.5 Spatial patterns of insect abundance and vegetation features along the 900-m pitfall transect with a 10-m trap spacing (91 total). (A) Vegetative cover determined from eight point samples around each pitfall station. (B) Number of captures per day of the darkling beetle Eleodes extricata in 1991, 1993, and 1994 (years that had three sample months; Fig. 10.1). (C) Distance to nearest shrub from each pitfall station (maximum, 10 m) with all shrubs included, and distance to nearest four-wing saltbush, Atriplex canescens. (D) Number of workers captured per day of the wood ant, Formica obscuripes, during 3 years. (E) Number of arid-land subterranean termites (Reticulitermes tibialis) recovered from toilet paper baits positioned on the soil surface at each pitfall–trap station. Values are the combined total from samples in June, July, and August 1993.

226

Ecology of the Shortgrass Steppe 0.6

(a) E. extricata abundance

0.4

91 93 94

Autocorrelation

0.2 0 -0.2 -0.4 0.5

91 93 94

(b) F. obscuripes abundance

0.3 0.1 -0.1 -0.3 -0.5 0

50

100

150

200

250

300

350

400

Lag Distance (m)

Figure 10.6 Autocorrelations of abundance of beetles, ants, termites across increasing distances between sample pairs (lag length). (A) Eleodes extricata. (B) Formica obscuripes.

and Wiens, 1995; McIntyre, 2000), but that the strength of the beetle–environment relationship differs across spatial scales (Bossenbroek et al., 2004; Hoffman and Wiens, 2004). Cross-correlation in F. obscuripes ants in pitfall traps showed strong positive correlations among years within the same traps (zero lag) and, to a lesser degree, at 30-m lags on each side of zero (Fig. 10.7C), likely a reflection of relative constancy in colony spacing during the study. Ant colony densities along three transects that span the same topographic gradient (Fig. 10.4B) reveal a shifting boundary in ant colony distribution as one moves horizontally across the pasture (Fig. 10.8A). This variation in ant colony distribution of the entire pasture was analyzed with anisotropic autocorrelation (Fig. 10.8B). The autocorrelation in colony density decreases along the north– south direction (0º), persists across greater lag lengths in the northeast–southwest (45º) and east–west directions (90º), and declines sharply in the southeast–northwest direction (135º), which most closely corresponds to the direction of the topographic gradient (Fig. 10.4B). Directions that run along the topographic contours (45º and 90º) exhibit positive autocorrelation with lags less than 700 m, whereas those that span topographic changes (0º and 135º) become negative at lags more than 400 m (Fig. 10.8B). The degree of change across lag lengths shown by anisotropic autocorrelations is therefore consistent with the position of the topographic gradient. Thus, the slope and aspect are closely tied to the density and distribution of harvester ants. This spatial coupling of ant density with topography, in turn, has important consequences for community interactions and ecosystem processes, as described next (see also Crist and Wiens, 1996; MacMahon et al., 2000).

Insect Populations, Community Interactions, and Ecosystem Processes 227

0.6

(a) E. extricata abundance among years

91 x 93 91 x 94 93 x 94

0.4 0.2 0 -0.2 -0.4 0.5 0.3

(b) E. extricata abundance and vegetation cover

0.1

Cross correlation

-0.1 -0.3

94 x Grass

94 x Bare ground

-0.5 0.8 0.6

91 x 93 91 x 94 93 x 94

(c) F. obscuripes abundance among years

0.4 0.2 0 -0.2 -200

-150

-100

-50

0

50

100

150

200

Lag Distance (m)

Figure 10.7 Cross-correlations between insect abundances across years. (A) Eleodes extricata between different pairs of years. (B) Formica obscuripes between different pairs of years. (C) Formica obscuripes between different pairs of years.

Community Interactions Patterns of distribution and abundance impinge on the strength and variability of species interactions within heterogeneous landscapes. For example, population densities and the activities of individual consumers affect the quantity of plant biomass removed by grasshoppers (Joern, 1987; Mitchell and Pfadt, 1974) or the rate at which seeds are harvested by ants (Crist and Wiens, 1994). The patterns of insect distributions in time and space therefore have important implications for community interactions in the shortgrass steppe. Plant–Herbivore Interactions Grasshoppers are the most important aboveground insect herbivores in rangeland ecosystems (Watts et al., 1982), often removing 20% to 25% of aboveground

Ecology of the Shortgrass Steppe

Colony density (# ha-1)

228

30

(A)

200 m 500 m 800 m

25 20 15 10 5 0 200

0

400

600

800

1000

1200 1400

1600

700

800

Distance (m)

0.8

Autocorrelation

0.6

(B)

0.4 0.2 0.0

0o 45o 90o o 135

-0.2 -0.4 -0.6

0

100

200

300

400

500

600

Lag Distance (m) Figure 10.8 (A) Colony density of Pogonomyrmex occidentalis across a topographic gradient from three different horizontal positions (see Fig. 10.4B). (B) Anisotropic autocorrelation of colony density from a grid of quadrat counts from the map of the entire pasture. The 0º and 135º directions run across the gradient, whereas the 45º and 90º directions run more parallel to contours.

primary productivity (Hewitt and Onsager, 1983; Mitchell and Pfadt, 1974). Plant consumption by grasshoppers is highly variable, however, and this variability in consumption may be expressed at spatial scales ranging from microhabitat to region and at temporal scales from days to decades (Joern, 1987). Differential effects of grasshopper herbivory on the plant community result from variability in feeding preferences of grasshopper assemblages. For acridids, diet preference and breadth are related to subfamily membership. In mixed-grass prairie, Joern (1983a) found that gomphocerines, which are principally grass feeders, used an average of 8.0 plant taxa in their diets; melanoplines, which feed mostly on forbs, used an average of 17.1 plant taxa. Diet breadth may vary in predictable ways among regions as well. Grasshoppers in the shortgrass steppe feed on fewer plants than in mixed-grass sites (Joern, 1987), possibly because of a decreased plant community dominance by B. gracilis as one moves from the shortgrass steppe to mixed-grass communities, which include codominants such as western

Insect Populations, Community Interactions, and Ecosystem Processes 229

wheatgrass (Agropyron smithii) and little bluestem (Schizachyrium scoparium). Despite the breadth of grasshopper diets, some plant species are consistently consumed in greater quantities than predicted by their relative abundance in the plant community (Mitchell, 1975; Mulkern, 1967). The effects of selective feeding by grasshoppers on plant community structure and ecosystem functioning have received considerably less attention (e.g., Rodell, 1977). Grasshoppers can clearly reduce their food supply, as shown by studies that demonstrate density-dependent regulation of population size (Belovsky and Joern, 1995; Belovsky and Slade, 1995; Kemp and Dennis, 1993). Because many grasshopper species that reach outbreak densities are oligophagous or polyphagous (Joern, 1987), grasshopper feeding may have widespread rather than selective effects within a plant community. Such predictions based on diet preferences could be used to predict the effects of grasshopper herbivory on plant community structure. In the shortgrass steppe, a significant alteration of plant community structure in response to grasshopper herbivory would likely involve changes in dominance by B. gracilis. Dyer and Bokhari (1976) found that laboratory-grown B. gracilis plants responded to grasshopper grazing by reallocating energy to belowground processes such as increased tiller production, respiration, and root exudation. This suggests that the effects of grasshopper grazing might be similar to those of livestock grazing in which there is an increased dominance by grazing-tolerant B. gracilis (Lauenroth and Milchunas, 1992; Milchunas et al., 1988, 1989); however, this hypothesis has not been tested in the field. The white grub larvae of June beetles (Phyllophaga fimbripes and P. crinata) are belowground herbivores that cause considerable mortality of perennial grasses. White grubs may exceed densities of 50/m2 and have substantial impacts on plant communities in rangelands (Ueckert, 1979; Watts et al., 1982; Wiener and Capinera, 1980). Root feeding by larvae produce killed patches of blue grama that are generally less than 0.1 ha in size but can reach 1 ha or more (Milchunas et al., 1990; Ueckert, 1979). Grub kill areas have altered plant and arthropod communities from those observed in surrounding areas. Rottman and Capinera (1983) found a higher plant diversity in grub kill compared with undamaged areas primarily as a result of an increased number of forb species immediately after grub disturbances (cf. Peters et al., chapter 6, this volume; Ueckert, 1979). Similarly, Milchunas et al. (1990) recorded decreased dominance of B. gracilis and Opuntia polyacantha in grub disturbances, resulting in higher plant diversity compared with surrounding areas. Overall arthropod diversity was relatively unchanged in grub kill areas, although some taxa showed differential effects to grub disturbance (Rottman and Capinera, 1983). For example, carabid (Harpalus spp.) and chrysomelid (Phyllotreta spp.) beetles were more abundant in grub disturbances than in surrounding areas; spiders, mites, and collembolans had reduced abundances in disturbances compared with undisturbed areas (Rottman and Capinera, 1983). Seed–Granivore Interactions Harvester ants (Pogonomyrmex spp.) are widespread and abundant insect granivores in rangelands (MacMahon et al., 2000). Studies on P. occidentalis have

230

Ecology of the Shortgrass Steppe

shown that foraging activity may significantly alter the density and distribution of seeds in the soil (Coffin and Lauenroth, 1990; Crist and MacMahon, 1992; MacMahon et al., 2000; Mull and MacMahon, 1996). Rogers (1974) estimated that 1% to 5% of the total seed pool in soil was harvested by ants at the CPER, which is lower than that found for P. occidentalis in other semiarid ecosystems (9% to 26% [Crist and MacMahon, 1992]). Ants selectively harvest seeds so that preferred species show greater losses from soil (Crist and MacMahon, 1992) and greater accumulation in ant nests (Coffin and Lauenroth, 1990) compared with surrounding areas. Measurements of native seed preferences correlate with ant effects on soil seeds (Crist and MacMahon, 1992), which suggests that seed-choice experiments provide a relative measure of the interaction strength between granivorous ants and various seed species. Crist and Wiens (1994) conducted seed-dish experiments to examine how seed removal by ants was influenced by seed species, plant patch structure, and cattle grazing. Broad-scale (pasture-level) differences in vegetation resulting from grazing and the presence of predators such as horned lizards affected both ant foraging patterns and colony activity, and comprised the largest component of variation in seed harvest. Within pastures, ants removed seeds more rapidly from patches of bare ground or grass than from cacti or shrubs. Within patches, seeds of annual forbs (Lepidium densiflorum) were removed far more rapidly than seeds of dominant grasses B. gracilis and Buchloë dactyloides (Crist and Wiens, 1994). Interestingly, the preference for L. densiflorum in seeddish trials is corroborated by a large number of L. densiflorum seeds recovered from ant nests (Coffin and Lauenroth, 1990). As with insect herbivory, the selective influences of insect granivory on the plant community are largely on subdominant plant species. Changes in the abundance of subdominant plants, however, can have substantial effects on the overall richness and diversity of shortgrass plant communities (Milchunas et al., 1990). In comparison with their effects on dominant grasses, the effects of insect herbivores or granivores on plant species diversity have received far less attention. An increasing emphasis on the management of biodiversity in range ecosystems (Hart, 2001; West, 1994) may renew interest in the effects of insect consumers on plant and arthropod diversity (e.g., Milchunas et al., 1990; Rottman and Capinera, 1983). Predator–Prey Interactions I consider two aspects of the roles of insects in aboveground predator–prey interactions: insects as predators and insects as prey. Together, these comprise the complex suite of feeding interactions that structure food webs. Insects and other arthropods are a major part of food webs in the shortgrass steppe (Lauenroth and Milchunas, 1992), but there are few studies of aboveground food web structure. Insects as Predators From studies of the IBP Grassland Biome project, the most abundant arthropod predators in the shortgrass steppe include wolf spiders (Lycosidae), hunting spiders

Insect Populations, Community Interactions, and Ecosystem Processes 231

(Gnaphosidae), robber flies (Asilidae), and ground beetles (Carabidae) (Lavigne and Campion, 1978; Lavigne and Kumar, 1974; Lavigne et al., 1971). Searches along 1.6-km transects during 1969 and 1970 showed that “sightable groups” such as lycosids and asilids had abundances roughly 10 times those of mantids (Mantidae), grasshopper wasps (Sphecidae), jumping spiders (Salticidae), or tiger beetles (Cicindelidae). Asilids numbers varied slightly (67–99) along transects, but did not differ among pastures (Lavigne et al., 1971). A total of 120 lycosid spiders were sighted in a lowland area, and 38 to 73 were found in three upland areas with different grazing intensities (Lavigne et al., 1971). In contrast, Weeks and Holtzer (2000) found that lycosids were significantly more abundant in uplands than in lowlands as measured by several pitfall trap arrays in each area. Gnaphosids had similar abundances in the two topographic positions (Weeks and Holtzer, 2000). Grasshoppers are a major part of the diets of wolf spiders and robber flies (Lavigne et al., 1971), but it is unclear whether spiders or robber flies can significantly reduce grasshopper densities in the shortgrass steppe (Capinera, 1987). Studies in similar systems suggest that arthropod predation on grasshoppers is primarily on early nymphal stages and that vertebrates (especially birds) may have greater effects on population sizes (Belovsky and Slade, 1993; Joern, 1986; Schmitz, 1993). However, spiders and robber flies also feed on other insects, such as leafhoppers and ants, and therefore may be important generalist predators in the shortgrass steppe. One of the most abundant carabid beetles at the CPER is H. desertus, which is an omnivorous species (Lavigne and Campion, 1978). In an ecosystem stress experiment, conducted during 1971 to 1974, H. desertus increased dramatically in response to water and nitrogen amendments. The response may have been the result of increased plant biomass, litter, or available prey (Lavigne and Campion, 1978), although microclimate changes also occurred in watered plots. Another carabid, Pasimachus elongatus, is a voracious predator capable of slicing large arthropods in half with its powerful jaws. McIntyre (1995) estimated densities of P. elongatus to be 1000 to 30,000 ∙ ha–1 in lowland areas with shrub cover and 30 to 200 ∙ ha–1 in upland grass sites. The high densities of P. elongatus and its ability to feed on a wide variety of insects suggest that it is also an important insect predator, especially in lowland swales. Insects as Prey The large abundances and body sizes of grasshoppers make them favored prey for numerous birds and mammals, including some of the most common vertebrates in the shortgrass steppe. Tenebrionid and carabid beetles are also large-bodied insects that form a major part of vertebrate diets. For example, Stapp (1996) found that Coleoptera (mostly scarabs, carabids, and tenebrionids) and Orthoptera (mostly grasshoppers and crickets) comprised 43% to 55% and 10% to 33%, respectively, of the seasonal diet of the grasshopper mouse (Onychomys leucogaster); the more omnivorous deer mouse (Peromyscus maniculatus) consumed 8% to 31% beetles and 0% to 18% grasshoppers and crickets. Likewise, grasshoppers and beetles comprise a significant fraction of the diets of insectivorous birds (Wiens and

232

Ecology of the Shortgrass Steppe

Dyer, 1975). These insects are also important in the diets of rare vertebrates as well. For example, the threatened Mountain Plover (Charadrius montanus) consumes tenebrionid beetles as a large part of its diet (F. Knopf, May 2005). Food Webs From the feeding relationships described earlier, one can sketch the positions of several insects in aboveground food webs. The herbivores are dominated by grasshoppers and leafhoppers (Andrews, 1979; Lauenroth and Milchunas, 1992; Lavigne and Kumar, 1974). The important arthropods in the predator trophic levels are robber flies and spiders (especially lycosids [Lavigne and Kumar, 1974]). Carabid beetles, such as P. elongatus and H. desertus, are generalist predators and omnivores, and likely feed on different trophic levels; their roles have been largely ignored in grassland food webs despite their high abundances (Lavigne and Campion, 1978; McIntyre, 1995). Experimental evidence from mixed-grass rangelands indicate that significant top-down effects on insect herbivores are more likely to be from vertebrate rather than arthropod predators (Belovsky and Slade, 1993; Joern, 1986; but see Schmitz, 1993; Schmitz et al., 2000); the critical experiments have yet to be conducted in the shortgrass steppe. Similarly, rodent removal experiments in other semiarid ecosystems suggest that carabid and tenebrionid beetle populations are strongly influenced by rodent predation (Parmenter and MacMahon, 1988a, b). Some evidence suggests that predation by rodents can also influence the distribution and abundance of tenebrionid beetles in the shortgrass steppe (Stapp, 1997). Links to belowground food webs are also evident among the herbivores, granivores, and detritivores. Harvester ants retrieve large quantities of litter and feces (Rogers, 1974) that decompose in nests and enter the belowground food web—a process that may be facilitated by late-season termite activity in ant nests (Crist and Friese, 1994). Tenebrionid beetles likewise form a connection to belowground food webs because they feed mostly on aboveground detritus that is processed and decomposed in soil (Lavigne et al., 1971). Belowground food webs also have important effects on soils and plant productivity (Moore et al., chapter 11, this volume) and, in turn, aboveground feeding interactions. Food web studies are needed on the multichannel feeding linkages that include aboveground herbivores and soil detritivores in grasslands and terrestrial ecosystems in general (Polis and Strong, 1996).

Ecosystem Processes Insects are significant in the processing of energy, in soil modification, and in disturbance and patch dynamics in the shortgrass steppe (Lauenroth and Milchunas, 1992). These processes operate at different spatial and temporal scales, and are closely tied to population dynamics and community interactions. Energy and Material Flows Arthropods comprise less than 1% of the total heterotrophic biomass in the shortgrass steppe. Between 96% to 99% of the biomass in heterotrophs is estimated to

Insect Populations, Community Interactions, and Ecosystem Processes 233

occur in the belowground microflora (Lauenroth and Milchunas, 1992). Among arthropods, most biomass (84%) and energy flow (79%) occur below ground (Lauenroth and Milchunas, 1992). However, the fraction of the ANPP consumed by arthropods (17% with cattle present, 79% without cattle) is much larger than the fraction of NPP consumed by belowground arthropods (5%). Although most biomass and energy flow occur belowground, arthropods have a more important role as direct regulators of aboveground production (e.g., grasshoppers). Except for root feeders such as white grubs, arthropods primarily have an indirect effect on belowground production through the decomposition of organic matter. Therefore, from a whole-system view, aboveground arthropod energetics are primarily driven by biophagic feeding (consumption of live biomass) and belowground energetics are mostly saprophagic (consumption of dead biomass) in the shortgrass steppe (Lauenroth and Milchunas, 1992). Beyond this generalization, there are three poorly known but potentially important aspects of the roles of insects in energy and nutrient flows. First, insect groups such as ground-dwelling beetles and ants have both above- and belowground activities that may accelerate the rate at which aboveground organic matter becomes incorporated into soil. Second, the redistribution and processing of materials by animals are spatially heterogeneous, which contributes to the patch dynamics of energy and nutrients within ecosystems (Whicker and Detling, 1988). Finally, the redistribution and processing of materials differ among insect groups, depending on their life history, vagility, and population size and distribution. Although it would be impractical to examine these differences on a species-byspecies basis, one can focus on keystone species or representatives of functional groups (Lawton and Brown, 1994). Here I contrast the roles of insect species to illustrate how material redistribution and processing might vary among insect groups and across different spatial scales. Fine-Scale Processes Fine-scale patchiness is most obvious at the level of the individual plant. In the shortgrass steppe, B. gracilis accounts for around 90% of the total basal cover, which ranges from 20% to 40% (Milchunas et al., 1989). Plants have average basal areas of 394 cm2 separated by bare ground (Aguilera and Lauenroth, 1993). Individual B. gracilis plants tend to form small hummocks surrounded by bareground depressions. This microtopography affects spatial patterns of organic matter and nutrient accumulation, which are concentrated in plant patches (Burke et al., chapter 13, this volume; Hook et al., 1991). Patches of cacti, shrubs, and forbs also form islands of nutrient and organic matter, but these patch types are more widely spaced among the dominant B. gracilis. Fine-scale patchiness of individual plants impinges on the spatial patterns of insect redistribution and processing of materials: Movement patterns of insects among plant patches affect redistribution, and residency of insects within patches affects the quantity of material processed. The fine-scale patterns of insect movement and distribution described earlier are germane to these processes. For grasshoppers, patch residency is related to dietary preferences or thermoregulation (Anderson et al., 1979; Joern, 1987), and

234

Ecology of the Shortgrass Steppe

movements are related to body size or the spatial distribution of microhabitats (Joern, 1983b; With and Crist, 1995). Because grasshopper grazing can significantly reduce aboveground biomass, this reduction should vary spatially according to patterns of movement and microhabitat use. Patch-specific grasshopper feeding affects fine-scale material flows in three ways: assimilation of plant material that may leave the patch when grasshoppers move, unassimilated feces that are deposited in the feeding patch or in other patches, and uneaten clippings that remain within the patch. An estimated 75% to 85% of the plant biomass destroyed by grasshoppers goes to uneaten clippings and feces (Mitchell, 1975; Mitchell and Pfadt, 1974). Thus, grasshoppers are much more important in the processing of plants to litter than in energy storage and trophic conversion (Mitchell and Pfadt, 1974). Patterns of grasshopper movements and feeding among plant patches could substantially contribute to fine-scale heterogeneity in nutrients and organic matter. Tenebrionid beetles also differ in their microhabitat use among plant cover types. Movement rates are greatest in bare ground and grass cover, and residence times are higher in cactus than in shrub patches (Crist and Wiens, 1995; Crist et al., 1992). Because tenebrionids feed primarily on plant litter and seeds (Allsopp, 1980), the net effect of tenebrionid beetles should be to redistribute litter (and seeds) from bare ground and grass to cactus and shrub patches. The rate of consumption and redistribution of litter have not been measured, but they could be significant because tenebrionid populations have a biomass that is comparable with grasshoppers (Lauenroth and Milchunas, 1992). Tenebrionids also spend a significant amount of time in burrows or soil crevices, so that some of the redistribution occurs belowground. Dung beetles (Scarabaeidae) are another group that are important to fine-scale redistribution of nutrients, especially in grazed systems where fecal production by livestock is an important nutrient pathway (Lauenroth and Milchunas, 1992). The ball-rolling dung beetle, Canthon pilularius, is a seasonally common species in the shortgrass steppe that specializes on cattle feces. Mated pairs of beetles roll dung into large spheres and bury them several centimeters deep into the soil (Guertin, 1993). Their larvae then feed on these dung balls and pupate to the surface. Adult activity occurs primarily under warm, wet conditions in June and July (Guertin, 1993), so that belowground transport of feces by dung beetles likely occurs in distinct seasonal pulses. Harvester ants retrieve seeds, litter, and insects from foraging areas and accumulate these organic materials in their nests. Some of these materials collected by ants may also be deposited in refuse piles on, or in areas away from, nests. Seeds are most likely to be removed from bare-ground areas (Crist and Wiens, 1994), but because ants are selective in both removing and discarding seeds, patterns of seed redistribution are dissimilar to that produced by wind transport (T. O. Crist, unpublished data). Harvester ants also clip aboveground vegetation immediately surrounding the nest to create a cleared disk. This clearing may result in a substantial number of plants removed from disks (Clark and Comanor, 1975), and the resulting litter becomes incorporated into soil within or near ant nests. Although we lack estimates on the amount of organic matter transferred into nests, the

Insect Populations, Community Interactions, and Ecosystem Processes 235

cumulative effects of concentrating organic matter into nests are clearly evident (see “Roles in Soils,” this chapter). Arid-land subterranean termites, R. tibialis, are another social insect that may have considerable importance in the belowground transfer of organic matter, especially woody litter. The considerable abundance of termites in areas with woody Atriplex shrubs (Fig. 10.5E) suggests that this species deserves further study regarding the importance of its role in the nutrient dynamics of shrub patches. Broad-Scale Processes Material flows at the level of the plant patch are closely linked to insect movement and microhabitat use. Variability in the flows of energy and nutrients across topographic gradients, soils, and plant assemblages should be more closely linked to the population distributions of insects. Insects often exhibit localized movements, so that long-distance transport of materials by insects is probably less important than patch-level movement. Heterogeneity in the effects of insects on energy and nutrient flows across landscapes should therefore stem primarily from variation in population abundance. Some of the broad-scale variation in insect population abundance can be linked to topographic variation (Fig. 10.3). If ecosystem effects are related to insect abundance, then the detritus processing and movement by E. extricata should be highest on mid slopes. Similarly, the removal of seeds and processing of plant material by P. occidentalis is likely to be greatest in uplands (Crist and Wiens, 1996). In contrast, F. obscuripes uses woody debris in nest construction, preys extensively on arthropods, and tends aphids in shrubs; these patch-level processes should be more important in lowland areas where F. obscuripes is common. Reticulitermes tibialis is also more common in lowland areas, where patches of woody litter occur (Fig. 10.5). Past studies of broad-scale patterns of grasshopper abundance have focused primarily on grazing rather than on topographic variation (Capinera and Sechrist, 1982; Welch et al., 1991). Spatial variation in grasshopper abundances may be considerable, however, and at least some of this variation is the result of vegetation (Przybyszewski and Capinera, 1990). This suggests that grasshopper densities might also vary substantially across a toposequence along with changes in vegetation (e.g., Milchunas et al., 1989), and these spatial distribution patterns could translate into topographic variation in the rates of herbivory and litter production in grassland ecosystems. Roles in Soils The effects of Pogonomyrmex ants on the chemical and physical properties of soils have been studied in the shortgrass steppe (Rogers and Lavigne, 1974) and in several other grassland ecosystems (reviewed by MacMahon et al., 2000). Substantial enrichment of soil nitrogen and phosphorus occurs in nest sites compared with surrounding areas (Mandel and Sorensen, 1982; Rogers and Lavigne, 1974; Whitford and DiMarco, 1995). In the shortgrass steppe, an estimated

236

Ecology of the Shortgrass Steppe

Mound : Off-mound abundance

2.8 kg soil⋅ha–1 ∙ y –1 is moved from lower horizons and deposited on the surface (Rogers and Lavigne, 1974), as well as considerable redistribution and sorting of particle sizes (Mandel and Sorensen, 1982). Increased porosity and accumulation of organic matter result in a reduction in soil bulk density (Mandel and Sorensen, 1982; Rogers and Lavigne, 1974). Soil water may also be significantly higher in ant nests and disks compared with surrounding soils (Laundré, 1990). Biotic enrichment of soils in ant nests occurs in response to increased levels of water, nutrients, and organic matter. In a shrub steppe ecosystem, Friese and Allen (1993) found roots within P. occidentalis nests to have a greater spore density of mycorrhizal fungi compared with surrounding soils. At the CPER, levels of mycorrhizae were consistently higher in ant nests compared with surrounding soils across topography and grazing (Snyder et al., 2002). Several other functional groups of microorganisms were enriched in ant nests compared with surrounding soils (Friese et al., 1997). Soils sampled from ant nests and off-mound areas in the shrub steppe and shortgrass steppe showed greater numbers of colonyforming units of microorganisms in ant nests compared with surrounding soils in both environments where P. occidentalis occurred (Fig. 10.9). The heightened activity of decomposer functional groups suggests that enriched levels of organic matter decompose more rapidly in nests than in surrounding soils. In contrast to nest mounds, the cleared disk becomes depleted of organic matter, nutrients, and microorganisms. Microbial biomass and mineralization of carbon and nitrogen

4.0 3.5

Shortgrass steppe

3.0

Shrub steppe

2.5 2.0 1.5 1.0 Total Bacteria

Total Fungi

Chitin Cellulose Protein -----Decomposer groups -----

Figure 10.9 The relative increase in abundance (colony-forming units) of bacteria, fungi, and microbial functional groups in harvester ant mounds (Pogonomyrmex occidentalis) over nonmound areas in two rangeland ecosystems. The increase is expressed as the ratio of the means of mound and nonmound soil samples (three in each location). Off-mound samples were taken 3 m from nests in soils underneath the dominant plants at each site (blue grama [B. gracilis] in the shortgrass steppe and big sagebrush [Artemisia tridentata] in the shrub steppe).

Insect Populations, Community Interactions, and Ecosystem Processes 237

were lower in soil from the cleared disk compared with soil under surrounding plants (Kelly et al., 1996). Disturbance Agents Harvester ants have long been recognized as disturbance agents in rangelands (reviewed by Rogers, 1987), but the various components of the ant disturbance regime—area, frequency, and turnover rates—are only more recently recognized (Coffin and Lauenroth, 1988). Clearings around ant nests, fecal pat deposition by cattle, and pocket gopher mounds are responsible for most of the animal-induced disturbances in blue grama grassland (Peters et al., chapter 6, this volume). An estimated 2.5 m2⋅ha–1⋅y–1 are cleared by ants in high-density areas (Coffin and Lauenroth, 1988, 1990). Rates of colony mortality and establishment can increase with ant density (Wiernasz and Cole, 1995), however, so that the frequency and turnover rate of disturbance may be greater in high-density areas (as in upland plains) than in low-density areas (as in lowlands). Vegetation surrounding ant clearings can have greater cover and altered species composition compared with unaffected areas (Coffin and Lauenroth, 1990; Nowak et al., 1990). The patch dynamics of ant disturbances after nest abandonment are not well understood. Initially, after the cessation of aboveground clipping by ants, large numbers of annual plants are found, as well as scattered perennial grasses that resprout from roots (Coffin and Lauenroth, 1990). Plant colonization of abandoned ant mounds might differ from other more ephemeral animal disturbances such as pocket gophers (Carlson and Crist, 1999; Huntly and Inouye, 1988; Wu and Levin, 1994) because of long-term occupancy and alteration of nest sites (MacMahon et al., 2000; Whitford, 1997). Although increased water and nutrient availability may facilitate plant establishment, it is less clear how biotic enrichment affects the trajectory of patch dynamics. The greatly increased levels of mycorrhizal fungi, for example, could favor late-successional mycotrophic plants over ruderal nonmycotrophic species (Friese and Allen, 1993; Friese et al., 1997). White grubs impose a somewhat different disturbance regime than harvester ants. Grub kills substantially reduce plant cover and create openings for new plant establishment (Coffin et al., 1998; Milchunas et al., 1990; Rottman and Capinera, 1983; Ueckert, 1979), but soils are not substantially modified as with harvester ants. The size of disturbance can be considerably larger (up to 1 ha or more) than ant clearings, but the frequency is generally lower and in different locations within the landscape (Peters et al., chapter 6, this volume). Ant disturbances are more common in upland soils, whereas grub densities and disturbances are greater near cacti or shrubs where soils are more penetrable (Wiener and Capinera, 1980). On a regional scale, grub kill areas may affect 100 to 1000 ha of grassland (Watts et al., 1982; Wiener and Capinera, 1980). The appearance of grub kill areas is often associated with drought years (Watts et al., 1982), but it is unclear whether grub densities increase in dry years or whether grub-induced plant mortality is greater with drought stress. The distribution and dynamics of Phyllophaga populations are not well understood, and their importance in grassland disturbance may be underestimated (Watts et al., 1982).

238

Ecology of the Shortgrass Steppe

Linking Population, Community, and Ecosystem Processes Insect Populations and Ecosystem Processes I have focused on grasshoppers, beetles, and ants in the shortgrass steppe to illustrate how patterns of distribution and abundance have important ecosystem consequences. The flows of energy and nutrients among ecosystem compartments and the redistribution of materials by insects depend on their spatial distributions. Likewise, the frequency of disturbance and the roles in soils depend on insect density and distribution. I emphasized two spatial scales at which processes are likely to differ. First, the processing and transport of materials at the level of the plant patch is primarily a consequence of insect movements, patch residence times, and the dispersion of insects among patches. Second, the overall effects of a particular species or functional group on the flows of energy and nutrients vary with population abundance according to broad-scale effects of soils, topography, and grazing. Conceptually, at least two scales of heterogeneity are therefore important in considering the effects of insect species on ecosystem processes. Differences in insect processing or transport of materials among patch types may reinforce or decrease the fine-scale heterogeneity in carbon and nitrogen (Hook et al., 1991), depending on the net effect produced by the movement and patch residency of the insect. The same can be said for the movements of propagules such as seeds or spores. At broader scales, population-level differences in rates of consumption, litter production, and decomposition by insects should be considered in concert with variation in soil nutrients (carbon, nitrogen, and phosphorus) and plant cover (or productivity), which generally increase from uplands to lowlands (Clark and Woodmansee, 1992; Milchunas et al., 1989; Schimel et al., 1985). Furthermore, the boundary contrast in plant cover, soil nutrients, and organic matter between plant patches and bare ground is greater in uplands than in lowlands, so that the collective effects of insects on patch-level heterogeneity will depend on topographic position. Insect Species Diversity and Ecosystem Processes What is the role of insect species diversity in ecosystem processes in the shortgrass steppe? Although a diverse insect fauna is described in the shortgrass steppe (Kumar et al., 1976), few studies of insect diversity have been conducted in relation to ecosystem function. Kirchner (1977) measured arthropod numbers, species diversity, and guild structure in the ecosystem stress experiment at the CPER. Plots with water or water-plus-nitrogen amendments showed significant increases in overall arthropod biomass and numbers. Arthropod diversity was also initially higher in these plots, but treatments showed highly variable patterns of diversity in subsequent years. There were significant positive relationships between the change in diversity of plants and arthropods across years. Diversity of feeding guilds increased in watered treatments compared with nitrogen-amended and control plots as a result of increased numbers of arthropod predators relative to

Insect Populations, Community Interactions, and Ecosystem Processes 239

herbivorous guilds. Kirchner (1977) attributed some of the variability in diversity to unmeasured structural changes in the community. Further efforts are needed to link insect diversity with functional attributes of rangeland ecosystems (Watts et al., 1982; West, 1994). The relationship between species diversity and ecosystem function is a topic of ongoing interest and importance (Jones and Lawton, 1995; Lambers et al., 2004; Naeem and Li, 1997; Tilman et al., 1996). Some investigators focus on keystone species or ecosystem engineers that have effects in ecosystems disproportionately greater than that suggested by their biomass or abundance; others examine how species redundancy within functional groups affects ecosystem processes. The relative value of these and other species-based approaches to analyzing ecosystem function is likely to vary among systems or the particular function under investigation. In the shortgrass steppe, for example, we might view white grubs or harvester ants as important disturbance agents that have no functional equivalent (cf. MacMahon et al., 2000; Whitford, 1997). In contrast, ant species with smaller and less conspicuous nests have a greater role in scavenging and nutrient redistribution (Bestelmeyer and Wiens, 2003). Grasshopper effects on primary production or plant community structure might be viewed as functionally redundant, where several species can have similar feeding preferences (e.g., within a subfamily) and where some species deletion might not influence overall rates of herbivory. Similarly, tenebrionid beetles are primarily detritivores, and several abundant species can have substitutable roles in the processing of plant litter.

Spatial Scaling of Insect Abundance Ecologists are increasingly concerned with questions of how small-scale processoriented studies can be extrapolated to broad-scale landscape patterns, or the degree to which broad-scale processes constrain those operating at fine scales (Wiens, 1989). Linear extrapolation across scales is often conducted without reference to the form of variability (King et al., 1991) or without the recognition that different processes operate at different scales (i.e., processes are scale dependent [Wiens, 1989]). A variety of approaches are used to measure scale dependence in patterns and processes (Milne, 1991; Turner et al., 1991). One approach to measuring scaledependent patterns includes the analysis of scaling exponents derived from log–log transformations (Johnson et al., 1992a; Milne, 1991; Schneider, 1994). Here I provide an example of scaling exponents using colony densities of harvester ants, which differ in their spatial distribution across spatial scales. Semivariance measures the variation among points separated by different distances (lag length). A log–log regression of semivariance against lag length can be used to detect scales of spatial dependence. A linear relationship (on a log–log scale) in spatial dependence often occurs over a limited range of scales. As the lag length increases, however, the slope of this relationship may change as processes operating at broader scales produce a different form of spatial dependence. The fractal dimension, D, of the semivariance relationship is derived from the slope, m, of the regression line by D = (4 – m)/2, and can be used to index shifts in

240

Ecology of the Shortgrass Steppe 2.1 D = 1.64

log Semivariance

1.9 1.7 1.5 D = 1.80

1.3 1.1 0.9 0.7 3.5

4.0

4.5

5.0

5.5

6.0

6.5

7.0

log Lag Length (m)

Figure 10.10 Semivariance in colony densities of harvester ants obtained from grid counts of the mapped area in Figure 10.4B. The log–log relationship between semivariance and lag length indicates two distinct regions of variability in colony density, which were determined by piecewise linear regression. The fractal dimension, D, is determined from the slope, m, of the regression line by D = (4 – m)/2.

spatial patterning across scales (Milne, 1991). If semivariance is constant across lag length (random noise), then m = 0 and D = 2. The semivariance of colony densities of harvester ants (Fig. 10.4B) was calculated by overlaying a 16 × 32 grid of 50 × 50-m (0.25-ha) cells (total, 128 ha). A log–log plot of semivariance with lag length shows two distinct regions that were determined by piecewise linear regression (Fig. 10.10). The first region, from 50 to 450 m, has a fractal dimension of D = 1.80, which indicates a more homogenous distribution of colony density at scales less than 450 m. The second region, from 450 to 900 m, shows greater heterogeneity in colony density with a shift to D = 1.64. Broad-scale patterns of ant distribution can therefore be described by two domains of scale: one that is controlled by local variation within a topographic position and another that is controlled by broad-scale changes in topography (Fig. 10.10). A more detailed spatial analysis suggests that ant distributions are influenced by colony establishment and interactions at a local scale, and changes in soils and grazing regime at broader scales (Crist and Wiens, 1996).

Conclusions and Future Directions A great deal is known about the common insects in the shortgrass steppe. A large number of species have been recorded, the major contributors to overall biomass quantified, and the population dynamics and distribution of some species have been measured. A considerable amount is also known about the impacts of cattle grazing

Insect Populations, Community Interactions, and Ecosystem Processes 241

on insect communities. However, there are still major gaps in our understanding of insects in shortgrass steppe ecosystems. We lack information on the long-term dynamics of important insect species. Few data are available on broad-scale spatial distributions of insects. There is only sparse information on the structure of aboveground food webs. There is also very little known about the relationships between insect diversity and ecosystem functioning. Nonetheless, the knowledge base of past studies will make possible future advances in these areas. Because of the important relationships among soils, nutrient dynamics, and plant community structure, future linkages of insect population distribution with ecosystem processes would be facilitated if measurements are taken along topographic gradients. Studies of insects in the shortgrass steppe can serve as a model for understanding animal roles in other terrestrial ecosystems. First, heterotrophs have important roles in grasslands because of the large fraction of consumable biomass and relatively rapid turnover of organic matter (Lauenroth and Milchunas, 1992). Second, most aboveground insects can be readily observed and measured in the low, open plant canopy in the shortgrass steppe. Lastly, their tremendous abundance and diversity make them good candidates to examine the various relationships among population, community, and ecosystem processes.

Acknowledgments My thinking about insects in the shortgrass steppe has been influenced by discussions with John Wiens, Carl Friese, Kim With, Paul Stapp, Nancy McIntyre, and Brandon Bestelmeyer. John Capinera, Mike Vanni, Deb Peters, Indy Burke, and Becky Riggle provided several valuable suggestions on earlier versions of the manuscript. I gratefully acknowledge research support from the NSF (BSR-8805829 and DEB-9207010 to J. A. Wiens) and the USDA (National Research Initiative competitive grant 95031420 to T. O. Crist and C. F. Friese).

References Aguilera, M. O., and W. K. Lauenroth. 1993. Neighborhood interactions in a natural population of the perennial bunchgrass, Bouteloua gracilis. Oecologia 94:595–602. Allsopp, P. G. 1980. The biology of false wireworms and their adults (soil-inhabiting Tenebrionidae) (Coleoptera): A review. Bulletin of Entomological Research 70: 343–379. Anderson, R. V., C. R. Tracy, and Z. Abramsky. 1979. Habitat selection in two species of short horned grasshoppers: The role of thermal and hydric stress. Oecologia 38:359–374. Andrews, R. M. 1979. A shortgrass prairie macroarthropod community: Organization, energetics and effects of grazing by cattle. U.S. IBP Grassland Biome technical report no. 208. Colorado State University, Fort Collins, Colo. Barnes, P. W., and A. T. Harrison. 1982. Species distribution and community organization in a Nebraska sandhills mixed prairie as influenced by plant/soil–water relationships. Oecologia 52:192–201. Belovsky, G. E., and A. Joern. 1995. The dominance of different regulating factors for rangeland grasshoppers, pp. 359–386. In: N. Cappuccino and P. W. Price (eds.), Population dynamics: New approaches and synthesis. Academic Press, San Diego, Calif.

242

Ecology of the Shortgrass Steppe

Belovsky, G. E., and J. B. Slade. 1993. The role of vertebrate and invertebrate predators in a grasshopper community. Oikos 68:193–201. Belovsky, G. E., and J. B. Slade. 1995. Dynamics of two Montana grasshopper populations: Relationships among weather, food abundance, and intraspecific competition. Oecologia 101:383–396. Bestelmeyer, B. T., and J. A. Wiens. 2001. Ant biodiversity in semiarid landscape mosaics: The consequences of grazing vs. natural heterogeneity. Ecological Applications 11:1123–1140. Bestelmeyer, B. T., and J. A. Wiens. 2003. Scavenging ant foraging behavior and variation in the scale and nutrient redistribution among semiarid grasslands. Journal of Arid Environments 53:373–386. Bossenbroek, J. M., H. H. Wagner, and J. A. Wiens. 2004. Taxon-dependent scaling: Beetles, birds, and vegetation at four North American grassland sites. Landscape Ecology 20:675–688. Capinera, J. L. 1987. Population ecology of rangeland grasshoppers, pp. 162–182. In: J. L. Capinera (ed.), Integrated pest management on rangeland. Westview Press, Boulder, Colo. Capinera, J. L., and D. R. Horton. 1989. Geographic variation in effects of weather on grasshopper infestation. Environmental Entomology 18:8–14. Capinera, J. L., and T. S. Sechrist. 1982. Grasshopper (Acrididae)–host plant associations: Response of grasshopper populations to cattle grazing intensity. Canadian Entomologist 114:1055–1062. Capinera, J. L., and D. C. Thompson. 1987. Dynamics and structure of grasshopper assemblages in shortgrass prairie. Canadian Entomologist 119:567–575. Cappuccino, N., and P. W. Price (eds.). 1995. Population dynamics: New approaches and synthesis. Academic Press, San Diego, Calif. Carlson, J. M., and T. O. Crist. 1999. Plant responses to pocket-gopher disturbances across pastures and topography. Journal of Range Management 52:637–645. Clark, W. H., and P. L. Comanor. 1975. Removal of annual plants from the desert ecosystem by Western harvester ants, Pogonomyrmex occidentalis. Environmental Entomology 4:52–56. Clark, F. E., and R. G. Woodmansee. 1992. Nutrient cycling, pp. 137–146. In: R. T. Coupland (ed.), Natural grasslands. Vol. 8A, Ecosystems of the world. Elsevier Science, Amsterdam. Coffin, D. P., and W. K. Lauenroth. 1988. The effects of disturbance size and frequency on a shortgrass plant community. Ecology 69:1609–1617. Coffin, D. P., and W. K. Lauenroth. 1990. Vegetation associated with nest sites of the Western harvester ant (Pogonomyrmex occidentalis) in a semiarid grassland. American Journal of Botany 76:53–58. Coffin, D. P., W. A. Laycock, and W. K. Lauenroth. 1998. Disturbance intensity and above- and belowground herbivory effects on long term (14y) recovery of a semiarid grassland. Plant Ecology 139:221–233. Crist, T. O. 1998. The spatial distribution of subterranean termites in shortgrass steppe: A geostatistical approach. Oecologia 114:410–416. Crist, T. O., and C. F. Friese. 1994. The use of ant nests by subterranean termites in two semiarid ecosystems. American Midland Naturalist 131:370–373. Crist, T. O., D. S. Guertin, J. A. Wiens, and B. T. Milne. 1992. Animal movement in heterogeneous landscapes: An experiment with Eleodes beetles in shortgrass prairie. Functional Ecology 6:536–544. Crist, T. O., and J. A. MacMahon. 1992. Harvester ant foraging and shrub-steppe seeds: Interactions of seed resources and seed use. Ecology 73:1768–1779.

Insect Populations, Community Interactions, and Ecosystem Processes 243 Crist, T. O., and J. A. Wiens. 1994. Scale effects of vegetation on forager movement and seed harvesting by ants. Oikos 69:37–46. Crist, T. O., and J. A. Wiens. 1995. Individual movements and estimation of population size in darkling beetles (Coleoptera: Tenebrionidae). Journal of Animal Ecology 64:733–746. Crist, T. O., and J. A. Wiens. 1996. The distribution of ant colonies in a semiarid landscape: Implications for community and ecosystem processes. Oikos 76:301–311. Dyer, M. I., and U. G. Bokhari. 1976. Plant–animal interactions: Studies of the effects of grasshopper grazing on blue grama grass. Ecology 57:762–772. Friese, C. F., and M. F. Allen. 1993. The interaction of harvester ants and vesicular– arbuscular mycorrhizal fungi in a patchy semiarid environment: The effects of mound structure on fungal dispersion and establishment. Functional Ecology 7:13–20. Friese, C. F., S. J. Morris, and M. F. Allen. 1997. Disturbance in natural ecosystems: Scaling from fungal diversity to ecosystem functioning, pp. 47–65. In: D. T. Wicklow and B. Söderstrom (eds.), The Mycota IV. Environmental and microbial relationships. Springer-Verlag, Berlin. Guertin, D. S. 1993. Trade-offs between feeding and reproduction in a ball-rolling dung beetle, Canthon pilularius (L.). PhD diss., Colorado State University, Fort Collins, Colo. Hart, R. H. 2001. Plant biodiversity on shortgrass steppe after 55 years of grazing of zero, light, moderate, or heavy cattle grazing. Plant Ecology 155:111–118. Hewitt, G. B., and J. A. Onsager. 1983. Control of grasshoppers on rangeland in the United States: A perspective. Journal of Range Management 36:202–207. Hoffman, A. L., and J. A. Wiens. 2004. Scaling of the tenebrionid beetle community and its environment on the Colorado shortgrass steppe. Ecology 85:629–636. Hook, P. B., I. C. Burke, and W. K. Lauenroth. 1991. Heterogeneity of soil and plant N and C associated with individual plants and openings in North American shortgrass steppe. Plant and Soil 138:247–256. Huntly, N., and R. Inouye. 1988. Pocket gophers in ecosystems: Patterns and mechanisms. BioScience 38:786–793. Joern, A. 1982. Vegetation structure and microhabitat selection in grasshoppers (Orthoptera: Acrididae). Southwestern Naturalist 27:197–209. Joern, A. 1983a. Host plant utilization by grasshoppers (Orthoptera: Acrididae) from a sandhills prairie. Journal of Range Management 36:793–797. Joern, A. 1983b. Small-scale displacements of grasshoppers (Orthoptera: Acrididae) within arid grasslands. Journal of the Kansas Entomological Society 56:131–139. Joern, A. 1986. Experimental study of avian predation on coexisting grasshopper populations (Orthoptera: Acrididae) in a sandhills grassland. Oikos 46:243–249. Joern, A. 1987. Behavioral responses underlying ecological patterns of resource use in rangeland grasshoppers, pp. 137–161. In: J. L. Capinera (ed.), Integrated pest management on rangeland. Westview Press, Boulder, Colo. Johnson, A. R., B. T. Milne, and J. A. Wiens. 1992a. Diffusion in fractal landscapes: Simulations and experimental studies of tenebrionid beetle movements. Ecology 73:1968–1983. Johnson, A. R., B. T. Milne, J. A. Wiens, and T. O. Crist. 1992b. Animal movements and population dynamics in heterogeneous landscapes. Landscape Ecology 7:63–75. Jones, C. G., and J. H. Lawton (eds.). 1995. Linking species and ecosystems. Chapman and Hall, New York. Keeler, K. H. 1993. Fifteen years of colony dynamics in Pogonomyrmex occidentalis, the Western harvester ant, in western Nebraska. Southwestern Naturalist 38:286–289.

244

Ecology of the Shortgrass Steppe

Kelly, R. H., I. C. Burke, and W. L. Lauenroth. 1996. Soil organic matter and nutrient availability responses to reduced plant inputs in shortgrass steppe. Ecology 77:2516–2527. Kemp, W. P. 1992. Rangeland grasshopper (Orthoptera: Acrididae) community structure: A working hypothesis. Environmental Entomology 21:461–470. Kemp, W. P., and B. Dennis. 1993. Density dependence in rangeland grasshoppers (Orthoptera: Acrididae). Oecologia 96:1–8. Kemp, W. P., S. J. Harvey, and K. M. O’Neill. 1990. Patterns of vegetation and grasshopper community composition. Oecologia 83:299–308. King, A. W., A. R. Johnson, and R. V. O’Neill. 1991. Transmutation and functional representation of heterogeneous landscapes. Landscape Ecology 5:239–253. Kirchner, T. B. 1977. The effects of resource enrichment on the diversity of plants and arthropods in a shortgrass prairie. Ecology 58:1334–1344. Kumar, R., R. J. Lavigne, J. E. Lloyd, and R. E. Pfadt. 1976. Insects of the Central Plains Experiment Range, Pawnee National Grassland. Agricultural Experiment Station, Laramie, Wyo. Lambers, J. H. R., W. S. Harpole, D. Tilman, J. Knops, and P. B. Reich. 2004. Mechanisms responsible for positive diversity–productivity relationships in Minnesota grasslands. Ecology Letters 7:661–668. Lauenroth, W. L., and D. G. Milchunas. 1992. Shortgrass steppe, pp. 183–226. In: R. T. Coupland (ed.), Natural grasslands. Vol. 8A, Ecosystems of the world. Elsevier Science, Amsterdam. Laundré, J. W. 1990. Soil moisture patterns below mounds of harvester ants. Journal of Range Management 43:10–12. Lavigne, R. J., and M. K. Campion. 1978. The effect of ecosystem stress on the abundance and biomass of carabidae (Coleoptera) on the shortgrass prairie. Environmental Entomology 7:88–92. Lavigne, R. J., and R. Kumar. 1974. Population densities and biomass of above-ground arthropods subjected to environmental stress treatments on the Pawnee site, 1973. U.S. IBP Grassland Biome, technical report no. 268. Colorado State University, Fort Collins, Colo. Lavigne, R. J., L. E. Rogers, and J. Chu. 1971. Data collected on the Pawnee site relating to Western harvester ant and insect predators and parasites, 1970. U.S. IBP Grassland Biome, technical report no. 107. Colorado State University, Fort Collins, Colo. Lawton, J. H., and V. K. Brown. 1994. Redundancy in ecosystems, pp. 255–270. In: E. D. Schulze and H. A. Mooney (eds.), Biodiversity and ecosystem function. Springer-Verlag, New York. Lockwood, J. A., and D. R. Lockwood. 1991. Rangeland grasshopper (Orthoptera: Acrididae) population dynamics: Insights from catastrophe theory. Environmental Entomology 20:970–980. MacMahon, J. A., J. F. Mull, and T. O. Crist. 2000. Harvester ants (Pogonomyrmex spp.): Their community and ecosystem influences. Annual Review of Ecology and Systematics 31:265–291. Mandel, R. D., and C. J. Sorensen. 1982. The role of the Western harvester ant (Pogonomyrmex occidentalis) in soil formation. Soil Science Society of America Journal 46:785–788. McIntyre, N. E. 1995. Methamidophos application effects on Pasimachus elongatus (Coleoptera: Carabidae): An update. Environmental Entomology 24:559–563.

Insect Populations, Community Interactions, and Ecosystem Processes 245 McIntyre, N. E. 1997. Scale-dependent habitat selection by the darkling beetle Eleodes hispilabris (Coleoptera: Tenebrionidae). American Midland Naturalist 138:230–235. McIntyre, N. E. 2000. Community structure of Eleodes beetles (Coleoptera: Tenebrionidae) in the shortgrass steppe: Scale-dependent uses of heterogeneity. Western North American Naturalist 60:1–15. Milchunas, D. G., W. K. Lauenroth, P. L. Chapman, and M. K. Kazempour. 1989. Plant communities in relation to grazing, topography, and precipitation in a semiarid grassland. Vegetatio 80:11–23. Milchunas, D. G., W. K. Lauenroth, P. L. Chapman, and M. K. Kazempour. 1990. Community attributes along a perturbation gradient in shortgrass steppe. Journal of Vegetation Science 1:375–384. Milchunas, D. G., O. E. Sala, and W. K. Lauenroth. 1988. A generalized model of the effects of grazing by large herbivores on grassland community structure. American Naturalist 132:87–106. Miller, R. H., and J. A. Onsager. 1991. Grasshopper (Orthoptera: Acrididae) and plant relationships under different grazing intensities. Environmental Entomology 20:807–814. Milne, B. T. 1991. Lessons from applying fractal models to landscape patterns, pp. 199–235. In: M. G. Turner and R. H. Gardner (eds.), Quantitative methods in landscape ecology. Springer-Verlag, New York. Mitchell, J. E. 1975. Variation in food preferences of three grasshopper species (Acrididae: Orthoptera) as a function of food availability. American Midland Naturalist 94:267–283. Mitchell, J. E., and R. E. Pfadt. 1974. A role of grasshoppers in a shortgrass prairie ecosystem. Environmental Entomology 3:358–360. Mulkern, G. B. 1967. Food selection by grasshoppers. Annual Review of Entomology 12:59–78. Mull, J. F., and J. A. MacMahon. 1996. Factors determining the spatial variability of seed densities in a shrub-steppe ecosystem: The role of harvester ants. Journal of Arid Environments 32:181–192. Naeem, S., and S. Li. 1997. Biodiversity enhances ecosystem reliability. Nature 390:507–509. Nowak, R. S., C. L. Nowak, T. DeRocher, T. Cole, and M. A. Jones. 1990. Prevalence of Oryzopsis hymenoides near harvester ant mounds: Indirect facilitation by ants. Oikos 58:190–198. Onsager, J. A. 1987. Integrated management of rangeland grasshoppers, pp. 196–204. In: J. L. Capinera (ed.), Integrated pest management on rangeland. Westview Press, Boulder, Colo. Parmenter, R. R., and J. A. MacMahon. 1988a. Factors influencing species composition and population sizes in a ground beetle community (Carabidae): Predation by rodents. Oikos 52:350–356. Parmenter, R. R., and J. A. MacMahon. 1988b. Factors limiting populations of arid-land darkling beetles (Coleoptera: Tenebrionidae): Predation by rodents. Environmental Entomology 17:280–286. Pfadt, R. E., and D. M. Hardy. 1987. A historical look at rangeland grasshoppers and the value of grasshopper control programs, pp. 183–185. In: J. L. Capinera (ed.), Integrated pest management on rangeland. Westview Press, Boulder, Colo. Polis, G. A., and D. R. Strong. 1996. Food web complexity and community dynamics. American Naturalist 147:813–846.

246

Ecology of the Shortgrass Steppe

Porter, S. D., and C. D. Jorgensen. 1988. Longevity of harvester ant colonies in southern Idaho. Journal of Range Management 41:104–107. Przybyszewski, J., and J. L. Capinera. 1990. Spatial and temporal patterns of grasshopper (Orthoptera: Acrididae) phenology and abundance in shortgrass prairie. Journal of the Kansas Entomological Society 63:405–413. Rodell, C. F. 1977. A grasshopper model for a grassland ecosystem. Ecology 58:227–245. Rogers, L. E. 1974. Foraging activity of the Western harvester ant in the shortgrass plains ecosystem. Environmental Entomology 3:420–424. Rogers, L. E. 1987. Ecology and management of harvester ants in the shortgrass plains, pp. 261–267. In: J. L. Capinera (ed.), Integrated pest management on rangeland. Westview Press, Boulder, Colo. Rogers, L. E., and R. J. Lavigne. 1974. Environmental effects of Western harvester ants on the shortgrass plains ecosystem. Environmental Entomology 3:994–997. Rottman, R. J., and J. L. Capinera. 1983. Effects of insect and cattle-induced perturbations on shortgrass prairie arthropod community. Journal of the Kansas Entomological Society 56:241–252. Schimel, D. S., M. A. Stillwell, and R. G. Woodmansee. 1985. Biogeochemistry of C, N, and P in a soil catena of the shortgrass steppe. Ecology 66:276–282. Schmitz, O. J. 1993. Trophic exploitation in grassland food chains: Simple models and a field experiment. Oecologia 93:327–335. Schmitz, O. J., P. A. Hamback, and A. P. Beckerman. 2000. Trophic cascades in terrestrial systems: A review of the effects of carnivore removals on plants. American Naturalist 155:141–153. Schneider, D. C. 1994. Quantitative ecology. Academic Press, San Diego, Calif. Snyder, S. R., T. O. Crist, and C. F. Friese. 2002. Variability in soil chemistry and arbuscular mycorrhizal fungi in harvester ant nests: The influence of topography, grazing and region. Biology and Fertility of Soils 35:406–413. Stapp, P. T. 1996. Determinants of habitat use and community structure of rodents in northern shortgrass steppe. PhD diss., Colorado State University, Fort Collins, Colo. Stapp, P. T. 1997. Microhabitat use and community structure of darkling beetles (Coleoptera: Tenebrionidae) in shortgrass prairie: Effects of season, shrub cover, and soil type. American Midland Naturalist 137:298–311. Tilman, D., D. Wedin, and J. Knops. 1996. Productivity and sustainability influenced by biodiversity in grassland ecosystems. Nature 379:718–720. Torell, L. A., and E. W. Huddleston. 1987. Factors affecting the economic threshold for control of rangeland grasshoppers, pp. 377–396. In: J. L. Capinera (ed.), Integrated pest management on rangeland. Westview Press, Boulder, Colo. Tscharntke, T., and H. J. Greiler. 1995. Insect communities, grasses, and grasslands. Annual Review of Entomology 40:535–558. Turchin, P. 1998. Quantitative analysis of movement. Sinauer Associates, Sunderland, Mass. Turchin, P., F. J. Odendaal, and M. D. Rausher. 1991. Quantifying insect movement in the field. Environmental Entomology 20:955–963. Turner, S. J., R. V. O’Neill, W. Conley, M. R. Conley, and H. C. Humphries. 1991. Pattern and scale: Statistics for landscape ecology, pp. 17–49. In: M. G. Turner and R. H. Gardner (eds.), Quantitative methods in landscape ecology. Springer-Verlag, New York. Ueckert, D. N. 1979. Impact of a white grub (Phyllophaga crinata) on a shortgrass community and evaluation of selected rehabilitation practices. Journal of Range Management 32:445–448.

Insect Populations, Community Interactions, and Ecosystem Processes 247 Uvarov, B. 1977. Grasshoppers and locusts: A handbook of general acrididology. Vol. 2. Centre for Overseas Pest Research, London, UK. Van Horn, D. 1970. Grasshopper population numbers and biomass dynamics on the Pawnee site from fall 1968 through 1970. U.S. IBP Grassland Biome, technical report no. 148. Colorado State University, Fort Collins, Colo. Watts, J. G., E. Huddleston, and J. C. Owens. 1982. Rangeland entomology. Annual Review of Entomology 27:283–311. Weeks, Jr., R. D. 1996. Habitat associations of ground-dwelling spiders (Arachnida, Araneae) on shortgrass steppe in northeastern Colorado: Implications for community organization and distribution. Masters thesis, Colorado State University, Fort Collins, Colo. Weeks, Jr., R. D., and T. O. Holtzer. 2000. Habitat and season in structuring grounddwelling spider communities (Araneae) in a shortgrass steppe ecosystem. Environmental Entomology 29:1164–1172. Welch, J. L., R. Redak, and B. C. Kondratieff. 1991. Effect of cattle grazing on the density and species of grasshoppers (Orthoptera: Acrididae) of the Central Plains Experimental Range, Colorado: A reassessment after two decades. Journal of the Kansas Entomological Society 64:337–343. West, N. E. 1994. Biodiversity of rangelands. Journal of Range Management 46: 2–13. Whicker, A. D., and J. K. Detling. 1988. Ecological consequences of prairie dog disturbances. BioScience 38:778–785. Whitford, W. G. 1997. The importance of biodiversity of soil biota in arid ecosystems. Biodiversity and Conservation 5:185–195. Whitford, W. G., and R. DiMarco. 1995. Variability in soils and vegetation associated with harvester ant (Pogonomyrmex rugosus) nests on a Chihuahuan Desert watershed. Biology and Fertility of Soils 20:169–173. Wiener, L. F., and J. L. Capinera. 1980. Preliminary study of the biology of the white grub Phyllophaga fi mbripes (LeConte) (Coleoptera: Scarabaeidae). Journal of the Kansas Entomological Society 53:701–710. Wiens, J. A. 1989. Spatial scaling in ecology. Functional Ecology 3:385–397. Wiens, J. A., T. O. Crist, and B. T. Milne. 1993a. On quantifying insect movements. Environmental Entomology 22:709–715. Wiens, J. A., N. C. Stenseth, B. Van Horne, and R. A. Ims. 1993b. Ecological mechanisms and landscape ecology. Oikos 66:369–380. Wiens, J. A., and M. I. Dyer. 1975. Rangeland avifaunas: Their composition, energetics, and role in the ecosystem, pp. 146–181. In: D. R. Smith (tech. coord.), Management of forest and range habitats for nongame birds. USDA Forest Service general technical report WO-1. Wiernasz, D. C., and B. J. Cole. 1995. Spatial distribution of Pogonomyrmex occidentalis: Recruitment, mortality, and overdispersion. Journal of Animal Ecology 64:519–527. With, K. A., and T. O. Crist. 1995. Critical thresholds in species’ responses to landscape structure. Ecology 76:2446–2459. Wu, J., and S. A. Levin. 1994. A spatial dynamic modeling approach to pattern and process in an annual grassland. Ecological Monographs 64:447–464.

11 Trophic Structure and Nutrient Dynamics of the Belowground Food Web within the Rhizosphere of the Shortgrass Steppe John C. Moore Jill Sipes Amanda A. Whittemore-Olson H. William Hunt Diana H. Wall Peter C. de Ruiter David C. Coleman

B

elowground organisms are key components of the trophic structure and they mediate the dynamics of nutrients of all terrestrial ecosystems. The interactions among assemblages of belowground microorganisms and their consumers mediate the cycling of plant-limiting nutrients, influence aboveground plant productivity, affect the course of plant community development, and affect the dynamic stability of aboveground communities following natural and anthropogenic disturbances (Clarholm, 1985; Ingham et al., 1985; Laakso and Setälä, 1999; Naeem et al., 1994; Tilman et al., 1996; Wall and Moore, 1999). The influence of belowground organisms on the aboveground plant community is heightened in systems such as the shortgrass steppe (Blair et al., 2000), given the relatively high percentage of plant production that is diverted belowground through plant roots. Many of the human-induced changes that the shortgrass steppe has been subjected to during the past 150 years fall outside the scope of the natural variations in climate and grazing. This conflict between the natural history of the shortgrass steppe and the more recent human legacy forms the backdrop of this chapter. First we present a detailed description of the belowground food web for the native shortgrass steppe and present its structure in terms of the patterns 248

Trophic Structure and Nutrient Dynamics 249

of trophic interactions, the distribution of biomass, the flow of energy, and the strengths of interactions. Second, we explore how three disturbances—managed grazing, agricultural practices, and climate change (altered precipitation and temperature, and elevated CO2)—have altered the structure of the belowground community. We conclude with a synthesis of the common patterns that we observed in the grassland’s response to these disturbances, and speculate on their consequences.

The Belowground Food Web Aboveground plant parts provide from 20% to 40% cover with exposed soil between them (Lauenroth and Milchunas, 1991). Much of the aboveground production remains in place as standing dead, rather than falling to the soil surface as litter. The ratio of shoot production to root production is roughly 1:1, contrasting sharply with forests, where far more production is allocated aboveground (Jackson et al., 1996; Milchunas and Lauenroth, 1993, 2000). Hence, in the shortgrass steppe, plant roots provide the major input of carbon to soil. As such, plant roots are the focal point of biological activity in soils (Coleman et al., 1983). The belowground food web of the shortgrass steppe consists of bacteria, fungi, protozoa, nematodes, arthropods, annelids, small mammals, and their interactions. Numerous studies of the soil biota of the shortgrass steppe have been conducted, dating back to the U.S. IBP of the late 1960s to early 1970s. During the IBP era, much of the work on soil biota cataloged the species diversity of various sites and began the process of determining population densities under native and manipulated conditions such as cattle grazing. The IBP studies of soil biota led to a body of work that linked various attributes of the soil (texture and structure), grass roots, nitrogen availability, and the feeding relationships among the taxa (Kumar et al., 1975; Coleman, 1976; Elliott and Coleman, 1977). These studies not only developed and refined the concepts of the rhizosphere, and the importance of soil biota to nutrient cycling and plant growth, but revealed three points that would change how we viewed communities and ecosystems. First, the studies used biomass estimates rather than densities, allowing for a common currency of carbon, nitrogen, or both. Second, the great diversity of soil biota made it necessary to develop a standard grouping scheme for taxa that shared similar attributes. Last, inorganic nitrogen was viewed as an integral component of the system, allowing it to be manipulated, modeled, and interpreted in the same vein as the soil biota. A Model Based on the Rhizosphere The food web diagrams and models we present are pictorial and mathematical representations of a conceptual model of the rhizosphere. Research of the shortgrass steppe led to the development of a widely used conceptual representation of

250

Ecology of the Shortgrass Steppe

the rhizosphere that emphasizes interaction (Coleman et al., 1983; Elliott, 1978; Moore et al., 2003; Trofymow and Coleman, 1982; Wall and Moore, 1999). From this perspective, the rhizosphere is defined as the narrow zone of soil that is influenced by plant roots and their products, and the trophic interactions that are affected by these products (Coleman et al., 1983; Moore et al., 2003; Trofymow and Coleman, 1982). Trofymow and Coleman (1982) proposed that a growing root could be viewed as a continuum of overlapping zones of activity, from the root tip to the crown, where different microbial populations have access to a continuous flow of organic substrates derived from the root. The first zone is the root tip. The tip is the site of root growth characterized by rapidly dividing cells of the rootcap and a low carbon-to-nitrogen ratio, or labile secretions or exudates that lubricate the tip as it passes through the soil. The exudates and sloughed root cells provide carbon for bacteria and fungi, which in turn immobilize nitrogen and phosphorus. Farther up the root is the region of nutrient exchange. This region is characterized by a higher carbon-to-nitrogen ratio or resistant products such as root hairs, and to a lesser extent the labile products of exudation. The birth and death of root hairs stimulate additional microbial growth (Bringhurst et al., 2001). The upper zones of the rhizosphere include regions of remineralization of nutrients by predators, symbiotic–mutualistic relations (mycorrhizae), and structure (Coleman et al., 1983; Moorman and Reeves, 1979; Wall and Moore, 1999). We present the connectedness, energy (nutrient) flow, and functional categories of descriptions that were proposed by Paine (1980) to describe the belowground food web of the shortgrass steppe. The connectedness food web describes the feeding relationships among groups. The energy flux food web describes the distribution of energy (usually in terms of biomass carbon or nitrogen) within and among the groups. The functional web describes the impacts that one group has upon the dynamics of the other groups.

The Connectedness Food Web Based on Functional Groups The connectedness description of the belowground food web of the shortgrass steppe is based on functional groups rather than species (Fig. 11.1) (Hunt et al., 1987; Moore et al., 1988). We forego a discussion of the major taxa, as they have been discussed at length in several venues (Moore et al., 1988; Wallwork, 1970; Walter and Proctor, 1999); however, representative genera are presented in Table 11.1. Functional groups are collections of species that are similar in terms of the following: (1) food sources, (2) feeding modes, (3) life history, and (4) habitat use. These criteria draw from niche theory (food, habitat, and time axes [sensu Schoener, 1974]) and from the dimensions of the differential equations that define the dynamics of the populations (mass, area, and time). The physiological attributes of the functional groups and the densities of each functional group are used to estimate feeding rates and interaction strengths, which are defined as the effects of one group’s dynamics on the dynamics of the other groups (Table 11.2).

Trophic Structure and Nutrient Dynamics 251

Shoots

Mycophagous prostigmata

Roots

Inorganic N

Labile substrates

Predatory Mites

Collembola Phytophagous nematodes

Nematophagous mites

Mycorrhizal fungi Cryptostigmata Saprophytic fungi

Resistant substrates

Predatory nematodes Mycophagous nematodes

Omnivorous nematodes

Flagellates Bacteria

Amoebae Bacteriophagous nematodes

Figure 11.1 The connectedness description of the belowground food web of the shortgrass steppe based on functional groupings of soil biota, plants, detritus, and soil nitrogen (Hunt et al., 1987; Moore et al., 1988).

The Energy Flow Food Web We have developed an energy flow description of the belowground food web, indexed by carbon flow (Fig. 11.2). The nutrient fluxes within the food web were estimated from the feeding rates between functional groups, the egestion rates of unassimilated consumption (feces, orts, and leavings), and the mineralization rates of carbon and nitrogen (metabolic release of CO2 and NH4+/NO3–). We present these formulations using units of carbon. Feeding rates are derived from population sizes and data on death rates and energy conversion efficiencies. The basic assumption underlying the calculation of feeding rates is that the annual (equilibrium) feeding rates should balance the annual death rate through natural death and predation (Hunt et al., 1987; O’Neill, 1969):

Fj =

d j Bj + Dj aj pj

(11.1)

where Fj is the feeding rate (measured in kilograms carbon per hectare per year ), dj is the specific death rate per year, Bj is the average annual (equilibrium)

252

Ecology of the Shortgrass Steppe

Table 11.1 Examples of Genera within the Functional Groups of the Shortgrass Steppe Belowground Food Web Functional Group

Description

Examples (Genera)

Predatory mites

Attack most soil invertebrates small enough to overcome

Nematophagous mites

Attack only nematodes

Predatory nematodes

Attack nematodes and bacteria (minimal) Consume bacteria and protozoa Feed on fungal cytoplasm Consume bacteria

Hypoaspis, Asca, Amblyseius, Rhodacarus, Gamasellodes, Macrocheles, Spinibdella, Cyta, Stigmaeus, Cocorhagidia Alliphis, Eviphis, Alycus, Alicorhagia, Ololaelaps, Veigaia Discolaimium, Mononchus

Omnivorous nematodes Fungivorous nematodes Bacteriophagous nematodes Collembola

Mycophagous prostigmata Cryptostigmata

Mesodiplogaster Aphelenchus, Aphelenchoides Acrobeloides, Pelodera, Rhabditis

Consume fungal hyphae and spores, algae, pollen Pierce fungal hyphae and consume fungal cytoplasm Consume fungal hyphae and spores

Amoebae

Consume bacteria

Flagellates

Consume bacteria

Phytophagous nematodes

Feed on plant roots

Folsomia, Isotoma, Isotomides Hypogastura, Tullbergia, Deuterosminthurus, Sminthurus Tydeus, Eupodes, Tarsonemus, Bakerdania, Pediculaster, Scutacarus, Speleorchestes Haplozetes, Passalozetes, Zygoribatula, Pilogalumna, Tectocepheus, Oppiella, Ceratozetes Acanthamoeba, Hartmanella, Tricamoeba, Mayorella, Varella Pleuromonas, Bodo, Mastigamoeba Helicotylenchus, Tylenchorhynchus, Xiphinema

From Kumar et al. (1975), Smolik and Dodd (1983), Hunt et al. (1987), and Moore et al. (1988).

population size (measured in kilograms carbon per hectare), Dj is the death rate resulting from predation (measured in kilograms carbon per hectare per year), aj is the assimilation efficiency, and pj is the production efficiency. For polyphagous predators, the feeding rate per prey type (Fij) is based on the relative abundances of the prey types and on prey preference:

Fij =

wij Bi n

∑ wkj Bk

Fj

(11.2)

k =1

where Fij is the feeding rate by predator j on prey i, and wij is the preference of predator j for prey i over its other prey types. The calculations of feeding rates

Trophic Structure and Nutrient Dynamics 253 Table 11.2 Estimates of the Constants Needed to Calculate Carbon and Nitrogen Fluxes, Fi (Eq. 11.1), Egestion Rates, Ei (Eq. 11.3) and Rates of Mineralization, Mi (Eq. 11.4) for Each Functional Group Presented in Figure 11.1

Functional Group

C:N

Turnover Rate per y

Assimilation Efficiency, %

Predatory mites Nematophagous mites Predatory nematodes Omnivorous nematodes Fungivorous nematodes Bacteriophagous nematodes Collembola Mycophagous prostigmata Cryptostigmata Amoebae Flagellates Phytophagous nematodes AM-mycorrhizal fungi Saprobic fungi Bacteria Detritus Roots

8 8 10 10 10 10 8 8 8 7 7 10 10 10 4 10 10

1.84 1.84 1.60 4.36 1.92 2.68 1.84 1.84 1.20 6.00 6.00 1.08 1.20 2.00 1.20 0.00 1.00

60 90 50 60 38 60 50 50 50 95 95 25 100 100 100 100 100

Production Efficiency, % 35 35 37 37 37 37 35 35 35 40 40 37 30 30 30 100 100

Biomass, kg C · ha–1 0.160 0.160 1.080 0.650 0.410 5.800 0.464 1.360 1.680 3.780 0.160 2.900 7.000 63.000 304.000 3000.000 300.000

AM, arbuscular mycorrhizae. From Hint et al., 1987.

started with the top predators, which suffer only from natural death, and proceeded working backward to the lowest trophic levels. The egestion rate (measured in kilograms carbon per hectare per year) is defined as the amount of the prey that is killed that is not assimilated by the predator per time step: Ej = (1 – aj)Fj

(11.3)

where Ej (measured in kilograms carbon per hectare per year) is the egestion rate by predator j of unassimilated prey and ai is the assimilation efficiency of predator j. The mineralization rate (measured in kilograms carbon per hectare per year) is the amount of the assimilated prey that is released in an inorganic form resulting from maintenance or access per time step: Mj = (1 – pj)ajFj

(11.4)

where Mj is the mineralization rate (measured in kilograms carbon per hectare per year) by predator j and pj is the production efficiency of predator j.

254

Ecology of the Shortgrass Steppe

Collembola

Shoots Phytophagous Nematodes Mycorrhizal fungi

Roots

Saprophytic fungi

Fauna

Cryptostigmata

Nematophagous mites

Mycophagous nematodes

Predatory nematodes

Flagellates N

Bacteria

Labile substrates

Predatory mites

Mycophagous prostigmata

Omnivorous nematodes Amoebae

Bacteriophagous nematodes

Resistant substrates

Figure 11.2 The energy flow description of the belowground food web presented in Figure 11.1. The boxes and arrows represent the relative biomasses of the functional groups and the estimates of the flow of energy (carbon) among functional groups (Eqs. 11.1 and 11.2, respectively). The flow to fauna represents all remaining flows not present in the diagram. (Adapted from Moore et al. [1988].)

The energy flux description reveals that the shortgrass steppe has a pyramidal structure that is compartmentalized into a root, bacterial, and fungal energy channel (Table 11.3) (Moore and Hunt, 1988). The pyramidal structure repeats itself within each energy channel. The majority of energy passes through the bacterial energy channel. The Functional Food Web We define the functional food web (Fig. 11.3) in terms of the interaction strengths. Interactions strengths are the elements of the community or Jacobian matrices (May, 1972, 1973), and represent the per capita—in this case, per biomass— effects of one group upon another at equilibrium. We estimated the interaction strengths from the population sizes and energy flow rates using Lotka-Volterra equations for the dynamics of the functional groups (de Ruiter et al., 1994; Moore et al., 1993): n dXi = Xi [ bi + ∑ cij Xj ] j =1 dt

(11.5)

Trophic Structure and Nutrient Dynamics 255 Table 11.3 Depiction of Energy Channels as Defined by the Proportion of Energy (Index by Carbon) Potentially Derived from Bacteria, Fungi and/or Plant Roots by the Different Functional Groups in the Belowground Food Web of the Shortgrass Steppe (adapted from Moore and Hunt [1988]) Energy Channel Functional Group

Bacteria

Fungi

Protozoa Flagellates Amoebae Ciliates

100 100 100

0 0 0

0 0 0

Nematodes Phytophagous nematodes Mycophagous nematodes Omnivorous nematodes Bacteriophagous nematodes Predatory nematodes

0 0 100 100 68.67

0 90 0 0 3.50

100 10 0 0 27.83

Microarthropods Collembola Cryptostigmata Mycophagous prostigmata Nematophagous mites

0 0 0 66.70

90 90 90 3.78

10 10 10 29.52

39.54

38.56

21.91

Predatory mites

Root

Estimates were based on the carbon fluxes obtained using Eqs. 11.1 and 11.2.

where Xi and Xj represent the population sizes of group i and j, respectively; bi is the specific rate of increase or decrease of group I; and cij is the coefficient of interaction between group i and group j. The matrix elements (αij) are defined as the partial derivatives near equilibrium: ␣ij = (∂

dX i dt

/ ∂X j )* . Values for the

interaction strengths are derived from the equilibrium descriptions by equating the death rate of group i resulting from predation by group j in equilibrium, cij Xi*Xj*, to the average annual feeding rate, Fij (Eq. 11.2) and the production rate of group j resulting from feeding on group i, cjiXj*Xi*, to aj Pj Fij. We assume that the long-term seasonal average population sizes of functional groups (Bi) approximate the theoretical steady-state densities (Xi*). Hence, the effect of predator j on prey i is ␣ij = cij X *j = −

Fij Bj

(11.6)

and the effect of prey i on predator j is ␣ij = cij X *j = −

aj pj Fij Bj

(11.7)

(A)

(B) Interaction Strength (yr -1 ) -20

Resource

-10 0 0.1 Top Predators

Nematophagous Mites Predaceous Nematodes Predaceous Nematodes Omnivorous Nematodes Ominivorous Nematodes Omnivorous Nematodes Amoebae Amoebae Bacteriophagous Nematodes Fungivorous Nematodes Mycophagous Prostigmata Cryptostigmata Collembola Bacteriophagous Nematodes Fungivorous Nematodes Bacteriophagous Nematodes Fungivorous Nematodes Flagellates Flagellates Flagellates Phytophagous Nematodes Phytophagous Nematodes Phytophagous Nematodes Bacteria Bacteria Bacteria Bacteria Bacteria Fungi Mycorrhizal Fungi Fungi Mycorrhizal Fungi Fungi Mycorrhizal Fungi Fungi Mycorrhizal Fungi Detritus Detritus Roots Roots

0.2 Consumer Predaceous mites Predaceous mites Nematophagous mites Predaceous mites Nematophagous mites Predaceous nematodes Predaceous nematodes Omnivorous nematodes Predaceous mites Predaceous mites Predaceous mites Predaceous mites Predaceous mites Nematophagous mites Nematophagous mites Predaceous nematodes Predaceous nematodes Predaceous nematodes Omnivorous nematodes Amoebae Predaceous mites Nematophagous mites Predaceous nematodes Predaceous nematodes Omnivorous nematodes Amoebae Bacteriophagous nematode Flagellates Fungivorous nematodes Fungivorous nematodes Mycophagous prostigmata Mycophagous prostigmata Cryptostigmata Cryptostigmata Collembola Collembola Bacteria Fungi Mycorrhizal fungi Phytophagous nematodes

Basal Resources 1

3

10-1 10 10 105

Feeding Rate (kg ha-1 yr -1 ) Figure 11.3 (A, B) The functional description of the belowground food web for the shortgrass steppe, Nunn, Colorado (de Ruiter et al., 1995; Moore et al., 1986). The bars correspond to the levels of feeding rates and interaction strengths for pairwise trophic interactions in a food web (see Fig. 11.1) arranged from resources (A) to consumers (B), starting with interactions at the base of the food web (basal resources) to the top of the food web (top predators). The estimates for the feeding rates were obtained using Eqs. 11.1 and 11.2. The pairwise interaction strengths represent the elements of the Jacobian matrix for the system of equations (Eq. 11.5) describing the trophic interactions among functional groups. The elements include the impact of a consumer on a resource (predator on prey), and the impact of the resource on the consumer (prey on predator), obtained from Eqs. 11.6 and 11.7, respectively. The asymmetry in the feeding rates (measured in kilograms nitrogen per

Trophic Structure and Nutrient Dynamics 257

Average Interaction Strength (ABS)

30 25 20 15 10 5 0 0

2

4 6 Number of prey per predator

8

10

Figure 11.4 The relationship between the effect of a predator on prey, y, presented as the average of the absolute values of the estimates obtained from Eq. 11.6, and the number of prey per predator, x, for fauna within the belowground food web (y = 14.1e –0.347x, r 2 = .679).

From the eigenvalues of the Jacobian matrix we can determine the dynamic stability of the community, and hence relate dynamic stability to the dynamics of nutrients. Stable Patterns of Interaction In the functional food web, we arranged the estimates of carbon flow and interaction strengths among the functional groups by trophic position (Fig. 11.3). Three sets of observations are evident. The first set deals with the distribution of biomass and feeding rates among functional groups with trophic position. The belowground food web of the shortgrass steppe forms a pyramid of biomass, with the majority of biomass positioned at the base of the food web and smaller amounts at the upper levels (Table 11.2). The feeding rates follow a similar pattern as the majority of nitrogen flows through the interactions at the base of the food web and declines with increased trophic position (Fig. 11.3). The second set of observations deals with the strengths of the trophic interactions. For the shortgrass steppe, the average strength of a predator’s impact on its prey follows a pattern that is important to stability (Fig. 11.4). May (1972) hypothesized that the average interaction strength between predator and prey

hectare per year) (A) and patterning of interaction strengths per year (B) within the Jacobian matrix with increased trophic level (bottom to top) were observed for food webs from the shortgrass steppe in Colorado (presented here), and agricultural sites in The Netherlands, Sweden, and Georgia, USA (de Ruiter et al., 1995; Moore et al., 1996).

258

Ecology of the Shortgrass Steppe

Stability (%)

100 80 60

unstable

40

stable

20 0 0.01 0.1 1 Life-like

0.01 0.1 1 Random

Figure 11.5 Monte Carlo trials comparing the lifelike food web of the shortgrass steppe with disturbed counterparts reveal that the asymmetrical patterning of interaction strength within the Jacobian matrix of the food web confers stability (de Ruiter et al., 1995). The black fraction in the bars denotes the percentage of stable matrices based on 1000 runs. For the lifelike matrices, the element values were sampled randomly from the uniform distributions [0, 2αij], in which αij is the element of the Jacobian matrix as derived from observations using Eq. 11.6 and Eq. 11.7. In the disturbed matrices, the values of the matrix elements were obtained by randomly permuting the nonzero pairs of elements after Yodzis (1981). The diagonal matrix elements referring to intragroup interference (αii) were set proportional at three levels of magnitude—0.01, 0.1, 1.0—to the specific death rates for that group.

should decrease as the number of prey species increased. Predators with few prey species would interact more strongly with each species than predators with many prey items. The third observation is that the two patterns discussed earlier—the patterning of nutrient flows and the patterning of interaction strengths with trophic position—are evident within the root, bacterial, and fungal energy channels (Fig. 11.5). Taken together, these observations lend strong support to the idea that the shortgrass steppe food web is compartmentalized, and that the compartmentalized structure is important to the stability of the ecosystem (May, 1972; Moore and de Ruiter, 1997; Moore and Hunt, 1988).

Conceptual Model of Community Structure, Nutrient Dynamics, and Stability These studies lead to the general conclusion that the energetic organization of communities forms the basis of ecosystem stability. We suggest that the changes in the trophic structure and material flow through the food webs brought about by disturbances should alter the distribution of biomass, the pattern of nutrient flow, and ultimately the patterning of interaction strength (Fig. 11.6). These changes in turn induce instabilities within the community much like those observed in the

Ratio of standardized interatction strengths

Trophic Structure and Nutrient Dynamics 259

100 10 1 0.1 0.01 0.001 0.50

Energy Channels: bacterial fungal root predator

1.00

1.50

2.00

2.50

3.00

3.50

4.00

4.50

Trophic position Figure 11.6 The asymmetry in the interaction strengths with increased trophic position for the belowground food web of the shortgrass steppe that was observed in Figure 11.3 can also be found within each energy channel (bacterial, fungal, and root). Each symbol represents the ratio of the pairwise interaction strengths between a consumer and a resource. Prior to taking the ratio, the interaction strengths were standardized by dividing each αij and αji by the average interaction strengths for all αij and αji, respectively. Because predators obtain a significant portion of their energy from more than one energy channel (Table 11.3), they were considered separately.

random webs (Fig. 11.5), or push the community away from its predisturbance steady state. A product of these periods of instability in an otherwise stable system (disturbance and recovery) are changes in material flow through the food web, and changes in the relationship between mineralization and immobilization of nitrogen. To illustrate these points we present the results of studies on the effects of three disturbances on the belowground food web: grazing by livestock, winter wheat agriculture, and elevated temperature and CO2. Impact of Grazing The plant communities of the shortgrass steppe evolved along with large grazing ungulates that included bison and pronghorn antelope, small mammals, arthropods, and nematodes (Milchunas et al., 1988; Milchunas et al., chapter 16, this volume). Although there has been some debate regarding whether grazing should be referred to as a disturbance (Milchunas and Lauenroth, 1993), we consider grazing to be a frequent and natural disturbance on the shortgrass steppe. The USDA Forest Service initiated a series of long-term studies in 1939 whereby exclosures were erected at the CPER to isolate areas from grazing by

260

Ecology of the Shortgrass Steppe

cattle. In 1991, the fences of several of the 1939 exclosures were moved to expose a portion of the exclosure to cattle. A new exclosure was built to prevent cattle grazing of an area that had been grazed each year since 1939. This new arrangement of fencing created four grazing treatments that combined historical treatments with new ones (historical/new): ungrazed, ungrazed/grazed, grazed, and grazed/ungrazed. Beginning in 1993, soil bacteria, saprophytic fungi, arbuscular mycorrhizal fungi, and protozoa were sampled at the end of July of each growing season through 1998. Nematodes were collected from 1992 through 1994. Long-term grazing had mixed effects on soil bacteria, saprobic fungi, protozoa, and nematodes. Early during the study, bacteria densities were higher in the historically nongrazed plots (ungrazed and ungrazed/grazed) than in the grazed plots (grazed and grazed/ungrazed). However, by 1994, no significant differences could be detected between long-term grazed and ungrazed plots. Long-term grazing had no significant effect on the densities of individuals of nematodes, propagules of fungi, or infection potential of arbuscular mycorrhizal fungi. Significant yearto-year variation in densities and higher densities under plants than between plants were observed. When viewed in a multivariate context, a different pattern emerged (Fig. 11.7). WallFreckman and Huang (1998) found that grazed plots (grazed and grazed/ungrazed) had

1994

Canonical variable 1

1993

1995

1997

Native/Native (N/N) Native/Grazed (N/G) Grazed/Grazed (G/G) Grazed/Native (G/N)

1996

YEAR: STATISTIC: 1993 1994 1995 1996 1997 Probability > F

Wilk’s λ .015 Pillai’s trace .124 Hotelling-Lawley trace .002 Roy’s greatest root 5 μg N⋅m⫺2 ⋅ hr⫺1) were most common during January and February (Fig. 14.1A) and were associated with snowmelt periods (80% of the high N2O fluxes during the winter occurred during periods of snowmelt). The winter N2O flux data show that significant fluxes occurred despite low soil temperature when microbial activity is assumed to be low. Because N2O may be produced from two processes—one aerobic (nitrification) and one

Exchange of Trace Gases in the Shortgrass Steppe 349

anaerobic (denitrification)—we deduce that the high fluxes are caused by denitrification during the wet periods associated with snowmelt. The four months of November through February typically represent 15% to 30% of the annual flux at each site. In some years, maximum mean monthly N2O flux rates were observed in January and February. The second peak in N2O emissions generally occurred during the May through September (Fig. 14.1A) period as a result of increased N turnover at the warmer temperatures, even though water is frequently limiting during the summer months. The high fluxes of N2O during the summer are primarily a result of nitrification because the observed soil water contents rarely are high enough to promote denitrification (Li et al., 1992; Parton et al., 1988, 1996). The large winter N2O emissions occurred only when soil, snow, and snowmelt fluctuations induced appropriate conditions. We attempted to quantify the effects of soil water and soil temperature on CH4 and N2O fluxes by aggregating data from all the treatments and sites we had studied on the SGS LTER. In the summer, at soil temperature more than 15 ºC, N2O fluxes tend to increase exponentially with increasing soil temperature (Fig. 14.2), and increase linearly as soil water-filled pore space (WFPS) increases from 10% to 50% (Fig. 14.3). We evaluated the soil temperature effect by using N2O flux data for cases when water was not limiting (WFPS was more than 30% and N2O fluxes were relatively high) and then fit an exponential curve to the data. We stratified the data into two groups: fertilized sites (Fig. 14.2A) and unfertilized sites (Fig. 14.2B). Similarly, we evaluated the effect of soil water on N2O flux by aggregating all the data from cases when soil temperature was not limiting (>15 ºC) and used a linear regression to represent the effect of WFPS (Fig. 14.3). The data were stratified into the same groups used for the temperature effect on N2O flux. The results (Figs. 14.2 and 14.3) show that N2O fluxes increase with increasing WFPS and soil temperature, that the fluxes are highest for the fertilized sites, and that they are similar for the unfertilized sites in sandy loam and swales. We evaluated the effect of soil temperature on CH4 uptake (Figs. 14.4 and 14.5) by aggregating the data for WFPS values from 15% to 25% (highest CH4 consumption) and calculating a linear regression of CH4 fluxes to soil temperature (Fig. 14.4). The data were further aggregated into two categories according to surface soil texture (Fig. 14.4). Although CH4 consumption increased linearly with increasing soil temperature (Fig. 14.4A) in the sandy sites (r 2 = .23), in the fine-textured sites soil temperature had little or no effect on CH4 consumption (r 2 = 0.0; Fig. 14.4B). The effect of WFPS on CH4 consumption was determined by using all the data with soil temperature more than 15 ºC (higher CH4 consumption) and then fitting a beta function to the data (Fig. 14.5). We used the beta function because CH4 consumption was highest for WFPS values from 15% to 25%, and decreased with WFPS values more than 25% or less than 15%. Again, we aggregated data into fertilized sites (Fig. 14.5A), nonfertilized sand (Fig. 14.5B), and nonfertilized clay sites (Fig. 14.5C). The results showed that peak CH4 consumption occurred at 15% WFPS for the sandy soils and 20% WFPS for the fine-textured soils. The unfertilized sandy soils had the highest CH4 consumption rates, followed by fertilized sandy sites and then by fine-textured sites (r 2 = .11, .2, .19, respectively).

350

Ecology of the Shortgrass Steppe

N2O (ug N m-2 hr-1)

20 15

a) Summer N2O WFPS > 0.3 Fertilized

10

y = 1.02e0.08x

5 0

(5)

N2O (ug N m-2 hr-1)

20 b) Summer N2O WFPS > 0.3 Nonfertilized

15

y = 0.97e0.04x

10 5 0

(5)

0

5

10

15

20

25

Temperature (C) Figure 14.2 Regression analysis of the effect of soil temperature on N2O flux at soil water-filled pore space (WFPS) more than 0.3 in fertilized sites (A) (n = 3; r 2 = .30) and unfertilized sites (B) (n = 3; r 2 = .11). Data points represented as triangles were considered outliers and were not included in the regression analysis.

Seasonal NO emissions appear to be dependent on temperature (r 2 = .36), with the highest fluxes occurring during the summer and the lowest during the winter (Fig. 14.6). It has been observed in both tropical forest and savanna ecosystems that the largest NO emissions are associated with precipitation events that follow long dry periods (Davidson et al., 1993; Johansson and Sanhueza, 1988). Spring and early summer were unusually wet in 1995, when May through July precipitation was double that received in 1994, which may account for the relatively small increases in flux after precipitation events in 1995. Zachariassen and Schimel (1991) measured CH4 fluxes on 8 days during four different parts of the growing season: green-up (mid June), peak biomass

Exchange of Trace Gases in the Shortgrass Steppe 351

35

a)

30

Fertilized

25 20 y = 0.03 + 14.51 x

15 10 5 0 -5

b)

30

Non-fertilized

25 20 15 y = 0.004 + 7.81 x

10 5 0 -5 0

0.1

0.2

0.3

0.4

0.5

0.6

WFPS Figure 14.3 Regression analysis of the effect of soil water-filled pore space (WFPS) on N2O flux when soil temperatures were more than 15 ºC in fertilized sites (A) (n = 3; r 2 = .23) and unfertilized sites (B) (n = 3; r 2 = .29). Data points represented as triangles were considered outliers and were not included in the regression analysis.

(late June), dry-down/senescence (early July), and postsenescence (mid to late September). The 24-hour mean flux rates observed for these periods were 0.004, –0.012, 0.007, and 0.034 μg N⋅m–2⋅s–1 respectively (Langford et al., 1992). Using these values, we estimated an annual plant-generated CH3 emission rate of 0.1 g N⋅m–2⋅y–1 for the shortgrass steppe. Impact of Landscape Position and Soil Texture We have evaluated the effects of soil texture and landscape position on trace gas fluxes using several sites (Mosier et al., 1998). Our results indicate that when landscape position is held constant, soil texture significantly affects both NOx and CH4 fluxes. Nitric oxide emissions and CH4 oxidation rates were higher in soils with the lowest clay content (Fig. 14.7). During a July through August sampling

352

Ecology of the Shortgrass Steppe

80

(A) 60

Sand

y = 28.34 + 0.73 x

40 20 0

(B) 60

Clay

y = 30.84 + 0.02 x

40 20 0 -5

0

5

10

15

20

25

30

Temperature (ºC) Figure 14.4 Regression analysis of the effect of soil temperature on CH4 uptake at soil water-filled pore space between 0.15 and 0.25 in coarse-textured soils (A) (n = 2; r 2 = .23) and finer textured soils (B) (n = 2; r 2 = .0).

period, NO emission from the undisturbed sandy loam soil was more than double that observed from the clay loam soil (P < .05; Fig. 14.7). At the same time, N2O emissions were lower from the soil having the lowest clay content. The effects of landscape position on trace gas fluxes of the shortgrass steppe are strong, but difficult to sort out from the direct effects of soil texture, as seen earlier. Our results from both sandy and clay catenas indicate that N2O fluxes from swales are typically higher than from mid slopes (e.g., on sandy catena, 54-month mean of 1.2 μg N⋅m–2⋅hr–1 for mid slopes and 2.0 μg N⋅m⫺2⋅hr⫺1 for swales [P < .05]; Fig. 14.8A). Conversely, CH4 uptake is higher in the summits and mid slopes than in swales for both sandy and clay catenas (36.7 μg C⋅m⫺2⋅hr⫺1 on sandy mid slopes compared with 21.3 μg N⋅m–2⋅hr–1 in the swales [P < .05]; Fig. 14.8B). Impact of Soil Nitrogen Additions Nitrous Oxide Annual N2O emissions from shortgrass steppe soils that have only natural N addition through atmospheric N deposition average about 100 g N⋅ha–1⋅y–1, which

Exchange of Trace Gases in the Shortgrass Steppe 353

100

(A)

80

Fertilized

60 y = 10

e

1-

0.05 x -0.15+ 0.05

(0 .05 x )

0 .15 + 0 .05

40 20 0

(B)

80

Non-fertilized sand

60 y = 15

e

1-

0.05 x -0.15+ 0.05

(0 .05 x )

0 .15 + 0 .05

40 20 0

(C)

80

Non-fertilized clay

60 y = 15

40

e

1-

0.05 x -0.20+ 0.05

(0 .05 x )

0 .20 + 0 .05

20 0 0

0.2

0.4

0.6

0.8

Soil Water-filled Pore Space (WFPS) Figure 14.5 Effect of soil water-filled pore space (WFPS) on CH4 uptake with soil temperature more than 15ºC in fertilized sites (A) (n = 2; r 2 = 0.11), unfertilized coarse-textured sites (B) (n = 2; r 2 = 0.20), and unfertilized fine-textured sites (C) (n = 2).

is 1.5% to 5% of the annual N input estimated from wet and dry deposition (Parton et al., 1988). We have conducted several fertilization studies to evaluate the effects of increased N availability on N2O emissions (Mosier and Schimel, 1991) by adding N to several landscape positions, using several different levels of fertilization. In studies simulating cattle urine deposition, we found that during the year after N application (Mosier and Parton, 1985; Mosier et al., 1982), N losses were small (0.5% to 1% of N added). In other studies with low N addition, we found a residual effect of fertilization on N2O emissions still evident after 12 to 13 years,

NO (ng N m-2 s-1)

60

(A)

PF PN

50 40 30 20 10 0

40

40

30

30

20

20

10

10

0

0 6/15/94

9/15/94 12/15/94 3/15/95

6/15/95

Soil temperature (ºC)

Precipitation (mm)

(B) 50

-10 9/15/95

Date Figure 14.6 (A) Fluxes of NO from pasture sites PF (native fertilized) and PN (native nonfertilized). (B) Precipitation and soil temperature measured at the CPER weather station (solid line), and soil temperature measured at the time of flux measurements (circles).

35 clay loam

30

sandy loam

25 20 15 10 5 0 NOx

N2O

CH4

Figure 14.7 Impact of soil texture on mean NOx and N2O emission rates, and CH4 uptake rates in a pasture with sandy loam soil (SL) and in a pasture with clay loam soil (CL). Gas flux data are means of weekly measurements over 2 years.

Exchange of Trace Gases in the Shortgrass Steppe 355 3.5

1991 1992 1993 1994

(A)

N2O Flux (ug N m-2 hr-1)

3 2.5 2

1

1.5 1 0.5 0

42

CH4 Uptake (ug C m-2 hr-1)

(B) 36 30 24 18 12 6 0 MN

MF

SN

SF

PN

PF

TC

MC

SC

Site Figure 14.8 (A) Mean annual N2O flux from nine sites at the CPER. (B) Mean annual CH4 uptake rates at nine sites in the CPER. Flux rates were calculated from mean monthly rates of weekly measurements from 1991 through 1994. Sand catena: TC, top clay; MF, midslope fertilized; MN, midslope-native; PF, pasture fertilized; PN, pasture native; SF, swale fertilized; SN, swale native. Clay catena: MC, midslope clay; SC, swale clay.

even though soil mineral N values were no longer elevated (Table 14.1, for sites that were fertilized in 1981 and 1982). Higher levels of N addition create a large response in N2O flux. Application of 22 kg N⋅ha–1⋅y–1 from 1976 through 1989 increased N2O emissions by an average of 73% during 1991 to 1994 (Figs. 14.1 and 14.8A). It is interesting to note

356

Ecology of the Shortgrass Steppe

that even though soil mineral N content was much higher in 1991 than in later years (an average of 10 μg N⋅g⫺1 of both NO3– and NH4+ in 1991 compared with an average of 1 μg N⋅g⫺1 of each in 1994), N2O fluxes were not lower in 1994. This observation suggests that it is the N turnover rate (coupled with precipitation patterns) that is controlling N2O emission, rather than bulk soil mineral N content. Bulk soil NH4+ concentration may not regulate nitrification rates (Davidson and Hackler, 1994), because the process may be limited by NH4+ supply on the microsite scale (Davidson et al., 1990). In the years immediately after fertilization, when bulk soil NH4+ and NO3– concentrations were high, NH4+ supply may have exceeded nitrifier demand. Methane Steudler et al. (1989) showed that mineral N additions to forest soils in the northeastern United States significantly decreased soil consumption of atmospheric CH4. Our work in the shortgrass steppe has shown that CH4 consumption decreases by 80% to 90% with the addition of 25 μg NH4+ N⋅g–1 soil and about 30% with the same amount of NO3– N (Bronson and Mosier, 1994). When we returned, after a number of years, to sites that had been fertilized in 1981, we were surprised to find that CH4 consumption rates were still much lower than at unfertilized locations (Mosier et al., 1991). Nitrogen Oxides (NOx) We measured NOx flux periodically from June 1994 through September 1995 to assess the effect of N fertilization last applied about 5 years earlier (Fig. 14.6A). During this 16-month period, the mean annual fluxes of NO (calculated by averaging across seasons for each observation location) were 24 and 11.5 μg NO N⋅m–2⋅hr–1 for fertilized and unfertilized sites, respectively. On an annual basis, about 0.21 and 0.10 g N⋅m–2⋅y–1 were emitted from these sites. Generally, fluxes at both sites followed the same temporal patterns, with emissions from the previously fertilized site exceeding those from the native site. Ammonia After the application of a urea solution, we measured NH3 volatilization continuously for 4 days, then rewet the soils and measured for another 4 days. Total NH3 volatilized during these periods was 1.5% and 14.1% of the N applied, respectively (Milchunas et al., 1988). Influence of Plant Species on Trace Gas Exchange We conducted a 2-year study to determine the impact of C3 and C4 plant composition on trace gas exchange (Epstein et al., 1998). We hypothesized that in the spring, during the time that C3 plants are most actively assimilating inorganic N, N gas fluxes would be lower within C3 communities than in C4 communities.

Exchange of Trace Gases in the Shortgrass Steppe 357

Conversely, we hypothesized that C4 communities would assimilate more N during the warmest part of the growing season, thus N gas emissions would be lower from C4-dominated communities than from C3-dominated ones during that time. We collected gas samples weekly from two sites differing in soil texture (one site with sandy clay loam and the other with clay), throughout two growing seasons. Plant functional type had no apparent influence on trace gas exchange in the clay soil. However, several differences were observed among plant communities on the sandy clay loam. We observed that CH4 uptake was greater (P < .05) on C4 plots than on C3 plots. Nitric oxide fluxes were significantly greater from C4 plots than from C3 plots during the drier of the 2 years of observation. The study indicated that under certain environmental conditions, particularly when factors such as moisture and temperature are not limiting, plant community composition can play an important role in regulating trace gas exchange. Impact of Land-Use Change in the Shortgrass Steppe Grasslands occupy large areas of the earth and likely influence the global biosphere. The soil–atmosphere exchange of trace gases within grasslands, and perturbations of this exchange by changing management and use of grasslands, has probably affected local, regional, and global atmospheres during the past century. Cultivation of grassland soils, for example, results in depletion of soil organic matter and reduces site fertility over time (Aguilar et al., 1988; Burke et al., chapter 13, this volume; Elliot and Cole, 1989; Ihori et al., 1995; Russel, 1929; Tieszen et al., 1982). Tillage increases both erosion and decomposition, leading to decreased soil organic matter and therefore a decrease in the ability of soil to retain mineral nutrients. These disturbances have long-term effects. Formerly cropped areas within the shortgrass steppe that were abandoned and have since reverted to grasslands appear to require several decades to return to typical native vegetation types and distribution (Coffin et al., 1996), and much longer to recover their soil organic matter (Burke et al., 1995). The most common agricultural conversion of Colorado grasslands has been to a dryland wheat-fallow cropping system (Hart, chapter 4, this volume). In this system, land is planted to wheat in alternate years and in the intervening years the land is not cropped but is kept weed free either by tillage or, in recent years, by herbicides. To study the effects of land use, we monitored gas fluxes from September 1992 through September 1995 in a wheat-fallow system (both phases), an undisturbed native pasture, a plowed pasture, and in an adjacent CRP site (Figs. 14.9 and 14.10). Methane uptake was always highest in the native pasture, relative to any of the other systems (Figs. 14.9A and 14.10a). Generally, the wheat-fallow sites exhibited CH4 consumption rates similar to those at the CRP site. However, CH4 uptake rates were generally lower in the recovering grassland (CRP) than in undisturbed grasslands (Figs. 14.9A and 14.10A). This may have been because soil moisture content was frequently higher in the CRP site during winter because of the inclusion of Agropyron cristatum in the plant community. Being taller than the native Bouteloua gracilis, this grass tends to collect more winter snow. During the 4 years of the study, CH4 uptake averaged 18.9 μg CH4

358

Ecology of the Shortgrass Steppe

50

(A) 40 30 20 10 0 30

(B)

WE WW

25

CRP

20

PN

15

PP

10 5 0 J

F

M

A M

J

J

A

S

O

N

D

Month Figure 14.9 (A) Mean monthly CH4 uptake rate. (B) Mean monthly N2O emission rate for five sites during the 3-year period September 1992 to September 1995. CRP, Conservation Reserve Program; PN, pasture native; PP, pasture plowed; WE, wheat east; WW, wheat west.

C⋅m–2⋅hr–1 at the CRP site, compared with 35.2 μg CH4 C⋅m–2⋅hr–1 at the native grassland site (Fig. 14.10A). Thus, the 5 to 8 years since the last cultivation of the CRP fields was apparently insufficient to return the microbial and physical processes responsible for CH4 oxidation to the native rate. Higher uptake rates in the CRP fields during the summer compared with those in wheat fields could be attributed to lower soil water content and lower N turnover rates, because the wheat fields were biannually fertilized with sewage sludge. The patterns in N2O emissions were generally opposite those of CH4 uptake, as we have observed before in grassland sites (Mosier et al., 1996). Thus, they were lowest in native fields and highest in plowed fields (Fig. 14.9B). All sites showed wintertime flux maxima in 1993, but in 1994 only the CRP site exhibited peak February flux. The following winter (1994–1995) was very warm and dry, thus eliminating the frozen soil N2O emission phenomenon frequently observed (Mosier et al., 1996). Nitrous oxide emissions were significantly lower in the

Exchange of Trace Gases in the Shortgrass Steppe 359

40

(A) 30

20

10

0

(B)

WE WW

15

CRP PN PP

10

5

0 1992-93

1993-94

1994-95

Mean

Year Figure 14.10 (A) Mean annual CH4 uptake rate. (B) Mean annual N2O emission rate for five sites during 1992 through 1995. CRP, Conservation Reserve Program; PN, pasture native; PP, pasture plowed; WE, wheat east; WW, wheat west. (Bars for the wheat-fallow fields are means of one complete cycle.)

native sites than in wheat or CRP sites (Fig. 14.10B). Reversion of the CRP to grassland did not lead to an observable decrease in N2O emissions after 8 years. Surprisingly, in the wheat sites, increased N2O emissions could not be linked with the timing of sewage sludge applications. Because the fields were merely harrowed after application of the sludge, much of the organic material remained on the soil surface, possibly allowing NH3 volatilization to continue as N was mineralized. Site management influences emissions as well, with a large N2O emission event occurring (Fig. 14.9B, site WE) coincident with combined surface sewage sludge application and rain.

360

Ecology of the Shortgrass Steppe

Effect of Time on the Return of Trace Gas Fluxes to Native Levels after Plowing To observe the changes in trace gas flux after plowing, we plowed a small area of native pasture and maintained the area free from plants for the following 2 years (Table 14.1, Fig. 14.10, site PP). Methane uptake rates declined to about 65% of the native site immediately after plowing and remained consistently lower throughout the following 3 years (Fig. 14.10A). The plowed site typically remained wetter through the year because the plot contained little vegetation. The combination of wetter soils and increased N cycling after plowing likely explain the reduced CH4 consumption (Mosier et al., 1996). Methane uptake rates were typically lowest during the winter and highest during the spring at both native and plowed sites (Fig. 14.9A). The uptake rates in the newly plowed site were generally higher than those in the wheat sites (Figs. 14.9A and 14.10A). Apparently, continued tillage and agronomic use tended to maintain CH4 oxidation at reduced rates. Nitrous oxide emissions increased immediately after plowing and continued to be higher for the next 3 years (Fig. 14.10B), although they did begin to decline markedly the third year after plowing (Fig. 14.10B). Unlike the native, wheat, and CRP fields, no large winter N2O emission events were observed at the newly plowed site. Because the major N and C sources that stimulate nitrification–denitrification are derived from decomposing plant material, after plowing the N2O release patterns tended to follow those of plant residue decomposition. The large increase in N2O after the single tillage event in this study was relatively short-lived. In 1994 and 1995, N2O fluxes were similar to those at the CRP site, with both sites having higher emissions in those years than the native grassland (Fig. 14.10B). To observe the long-term effect of reverting plowed fields to grasslands, we studied fields that were last cultivated 0 to 53 years before sampling and were allowed to revert back gradually to native vegetation (Fig. 14.11). During the 15-month measurement period (April 1992–July 1993), CH4 uptake (based on monthly means) averaged 36.0 and 36.3 μg⋅cm–2⋅hr–1, and N2O emissions averaged 1.9 and 1.3 μg N⋅m–2⋅hr–1 at native and old replanted sites, respectively. Neither CH4 nor N2O fluxes differed significantly between native and replanted sites (P < .05), and gas fluxes at both sites followed similar weekly and seasonal trends (Fig. 14.11). These results are an important indicator that 50 years is sufficient time for recovery of trace gas fluxes after cessation of cultivation. Burke et al. (1995; chapter 13, this volume) found that labile pools of organic matter had recovered on 50-year-old replanted fields, but that the total soil organic matter was still as much as 30% lower than that of never-cultivated sites. Trace gas fluxes follow the patterns of the labile soil pools. Effect of Elevated Carbon Dioxide on Trace Gas Exchange During the past few decades, the atmospheric concentration of CO2 has increased at historically unprecedented rates (IPCC, 2007). Increasing CO2 concentration has a direct effect on plant production and plant communities, but indirectly feeds back into a number of soil biotic systems that influence long-term ecosystem

Exchange of Trace Gases in the Shortgrass Steppe 361

10 9 8 7 6 5 4 3 2 1 0

N2O CH4

1995

1992

1987

1939

N1

N2

Last year tilled Figure 14.11 Mean annual CH4 uptake and N2O emission rates, calculated from 3 years of weekly flux measurements, for sites that were last tilled and allowed to revert gradually to native vegetation 0, 2, 7, and 53 years before flux measurements were made, and from two native grassland sites (N1 and N2).

viability (Hungate et al., 1997a,b,c; Owensby et al., 1993). These interactive feedbacks on the soil C and N cycles, and their influence on trace gas fluxes, have potentially important impacts on the global atmospheric budgets of these gases and on the long-term sustainability of the grassland. Our studies of trace gas fluxes in the shortgrass steppe demonstrate that such grasslands play an important role as consumers of atmospheric CH4, and producers of N2O (Mosier et al., 1991, 1996, 1997). The impact of elevated CO2 on the production and consumption of other trace gases (NOx, N2O, and CH4) is not well understood and has not yet been assessed in semiarid grasslands. The few short-term measurements of NOx, N2O, CH4, and CO2 fluxes in CO2 enrichment studies in other ecosystems give contradictory results, and long-term measurements had not been made within any ecosystem before the shortgrass steppe studies reported in Mosier et al. (2002a). To quantify the effects of doubling CO2 on shortgrass steppe ecosystem dynamics and trace gas exchange, we monitored the soil–atmosphere exchange of CO2, NOx, N2O, and CH4 weekly, year-round, April 1997 through November 2001, on treatments that included unchambered control, chambered with ambient CO2 (≈365 µmol⋅mol–1), and chambered with enriched CO2 (≈720 µmol⋅mol–1 CO2) plots in the Colorado shortgrass steppe (Lauenroth et al., chapter 12, this volume). Even though both C3 and C4 plant biomass increased under elevated CO2, and soil moisture content was typically higher than under ambient CO2 conditions,

362

Ecology of the Shortgrass Steppe

60

(A)

CH4

50

CO2

40 30 20 10 0 14

(B)

N 2O

12

NOx

10

H2O

8

Soil T

6 4 2 0 Ambient

Elevated

Figure 14.12 Effect of doubling atmospheric CO2 on trace gas exchange, soil water, and soil temperature. Data presented are averages of flux measurements, and soil water and temperature measurements made at each sampling day weekly between April 1997 and November 2001. Error bars represent the SD (n = 3). (A) Methane uptake and system respiration (CO2 flux). (B) Nitrous oxide and NOx flux, soil water, and temperature (T).

none of the trace gas fluxes were significantly altered by CO2 enrichment. During the 55 months of observation, NOx flux averaged 4.3 μg N⋅m–2⋅hr–1 in ambient CO2 And 3.1 μg N⋅m–2⋅hr–1 in enriched CO2 (Fig. 14.12A), and it was negatively correlated to plant biomass production. Under ambient and elevated CO2, N2O emission rates averaged 1.8 and 1.7 µg N⋅m–2⋅hr–1, CH4 flux rates averaged–31 and–34 μg C⋅m–2⋅hr–1, and ecosystem respiration averaged 43 and 44 mg C⋅m–2⋅hr–1, respectively (Fig. 14.12) (Mosier et al. 2002a,b). Methane oxidation tended to be higher in elevated CO2 than in control plots, whereas NOx and N2O tended to be lower, but not significantly so in either case.

Exchange of Trace Gases in the Shortgrass Steppe 363

Aboveground biomass production was higher under elevated CO2 (Morgan et al., 2001), and utilized more soil N (King et al., 2004). However, soil N mineralization was probably somewhat enhanced under elevated CO2 because of moister soils (Hungate et al., 1997a,b,c). The two opposing processes apparently offset each other, because NOx and N2O emissions, which reflect system N mineralization and nitrification, did not differ. Ecosystem respiration, which included soil and aboveground plant respiration, was not generally higher under elevated CO2. By analyzing the concentration of soil CO2 at different depths in the open-top chambers and calculating soil respiration, Pendall et al. (2003) found that elevated CO2 increased soil respiration by about 25% in a moist growing season and by about 85% in a dry season. Significant increases in soil respiration rates occurred only during dry periods. Some δ13C analyses of soil CO2 revealed that soil organic matter decomposition rates were more than doubled under elevated CO2, whereas rhizosphere respiration rates were not changed. Estimates of net ecosystem production, which account for both inputs and losses of C, suggest that soil C sequestration is not increased under elevated CO2 during dry years, but may be in wet years (Morgan et al., 2004; Pendall et al., 2004). Residual Effects of Carbon Dioxide We were interested to determine whether residual effects on microbial processes persisted after CO2 enrichment. The open-top chambers were removed at the end of the long-term study, and we conducted a short-term study to determine whether soil microbial processes were altered in soils that had been exposed to double-ambient CO2 concentrations during the previous five growing seasons. We measured the response of CO2, NOx, and N2O emissions and atmospheric CH4 uptake to water addition and water-plus-N fertilization to soils that had and had not (ambient) been exposed to elevated CO2. These responses are detailed in Mosier et al. (2003). Nitric oxide emissions from the ambient CO2 soils were higher than the elevated CO2 treatment when the soils were irrigated, and N2O emissions from elevated CO2 soils were significantly lower than the ambient CO2 soils (P < .05) when irrigated (Fig. 14.13A). This reflects the N-depleted state of the soils under elevated CO2 that occurs as a result of enhanced plant production (King et al., 2004). During the measurement period, NO-to-N2O ratios ranged between 2 and 5, with the highest in the elevated CO2 soils. When NH4NO3 was added, N2O flux tripled in elevated CO2 soils (P < .05), but increased only slightly in ambient CO2 soils (P > .05; Fig. 14.13B). Nitric oxide emissions and N2O emissions in N-fertilized soils increased markedly after each irrigation and precipitation event. Nitric oxide fluxes increased almost 10-fold with N addition in ambient CO2 soils, but only about fivefold in elevated CO2 soils. Nitric oxide emissions were significantly lower from elevated CO2 soils than from ambient CO2 soils, again indicating the N-depleted state of the elevated CO2 soils. Hungate et al. (1997a, c) found that, during wet-up, NO emissions were depressed by 55% in high nutrient conditions under elevated CO2 (ambient + 360 µmol⋅mol–1),

364

Ecology of the Shortgrass Steppe

70 60

(A)

50 40 30 20 10 0

(B) N2 O

160

NOx CH4

120

CO2

80 40 0 Ambient

Elevated

Figure 14.13 The effect of 5 years of CO2 enrichment on the flux of N2O, NOx, and CO2, and CH4 uptake after irrigation (A) and irrigation plus ammonium nitrate addition (10 g N⋅m–2) (B). Values are averages of 11 flux measurements made over a month-long observation period; error bars represent SD (n = 3).

whereas there was no difference among treatments in N2O emissions. They attributed the decreased NO emissions under elevated CO2 to increased N immobilization. Increased utilization of added N by soil microbes, thus a decrease in NO emissions, appears to be the case in this study as well. Plant growth during the time of the study was virtually nonexistent because of the very low amount of precipitation that had fallen the preceding year. Increases in ecosystem CO2 flux (dark chamber respiration, which includes plant, root, and soil microbial respiration) after water addition were similar in all soils. Only with the water-plus-N addition did CO2 fluxes from elevated CO2 soils exceed those from control and ambient soils (P < .05). Microbial respiration appears to be enhanced under elevated CO2 (Pendall et al., 2003), especially when microbes are not limited by water or N availability. Nitrogen addition appeared to stimulate

Exchange of Trace Gases in the Shortgrass Steppe 365

soil microbial respiration while decreasing NO emissions because of increased microbial immobilization of added N. After irrigation and irrigation plus fertilization, the rate of uptake of atmospheric CH4 was significantly greater (P < .05) in elevated CO2 soils than ambient CO2 soils (Fig. 14.13). Methane uptake rates were not measurably enhanced with N addition in ambient CO2 soils, but tended to be greater in elevated CO2 soils (P > .05). During the 5 years of CO2 enrichment, CH4 uptake rates tended to be higher under elevated CO2. This short-term study suggests that a microbial population developed under elevated CO2 capable of increasing utilization of atmospheric CH4. This is in contrast to the study by Ineson et al. (1998), who observed lower CH4 uptake rates under elevated CO2 within a free-atmosphere CO2 enrichment study in Switzerland. They also observed lower CO2 respiration rates and increased N2O emissions under elevated CO2. McLain et al. (2002) also observed lower CH4 consumption rates under elevated CO2 in a pine plantation. The decrease in CH4 consumption was attributed, in part, to wetter soils under elevated CO2. Soil conditions in the pine forest were likely much more comparable with the grassland soils in Switzerland (Ineson et al., 1998) than with the much drier conditions in the Colorado shortgrass steppe. The wetter soil conditions under elevated CO2 in the semiarid grassland likely produced more favorable conditions for methanotrophic activity, rather than limiting CH4 diffusion into the soil as in the Swiss grassland (Ineson et al., 1998) and in the pine forest (McLain et al., 2002). Hu et al. (2001) suggested that over the long term, soil microbial decomposition is slowed under elevated CO2 because of N limitation. Conversely, Hungate et al. (1997a) found that higher soil water content under elevated CO2 in an annual grassland stimulated soil N mineralization and resulted in greater plant N uptake. The N addition study confirms the observations of Pendall et al. (2003) that soil respiration is enhanced under elevated CO2 and N immobilization is increased, thereby decreasing NO emissions, despite the fact that we observed no general CO2-induced effect on NOx and N2O flux during the 5-year observation period (Mosier et al., 2002a,b).

Annual Nitrogen Trace Gas Losses for a Shortgrass Steppe Although cattle grazing contributes both to N redistribution within a pasture and potential losses through NH3 volatilization, the N loss rates are relatively small. Schimel et al. (1986) measured NH3 emissions from sets of microplots on midslope and swale positions of a sand catena in 1981 and 1982. They found that 27% of summer-applied N was lost through NH3 volatilization from the midslope soil during the month after application, whereas less than 1% was lost from swale soils during the same time period. They extrapolated NH3 losses observed in the direct measurements for the two catena positions to the pasture scale. Using seasonal rates of urine and feces decomposition, landscape position stratification data from Stillwell (1983), and measured NH3 rates, Schimel et al. (1986) estimated that 0.01 g N⋅m–2⋅y–1 was lost from the shortgrass steppe through NH3 volatilization. They assumed that pastures contained 70% upland (coarse-textured soil) and

366

Ecology of the Shortgrass Steppe

30% lowland (fine-textured soil). They also suggested that NH3 loss from senescing plant material is an important loss vector, emitting about 0.2 g N⋅m–2⋅y–1. Zachariassen and Schimel (1991) estimated from seasonal direct flux measurements that about 0.07 g NH3 N⋅m–2 was lost from a shortgrass pasture during this time The estimates of Schimel et al. (1986) were based on observed differences in total N content (1.5%) of peak standing biomass during the summer, N content of standing biomass (0.9% to 1.1%) in the late autumn, and estimated N translocation to roots in the autumn. Mosier and Parton (1985) and Parton et al. (1988) show that significant N losses through N2 emissions from denitrification are generally unlikely because the conditions for denitrification to occur in the system are rare and brief. More recent studies (Mosier et al., 1996, 1997) do suggest, however, that winter denitrification may result in significant N losses when conditions are conducive (frozen soil, snow accumulation, and brief warm temperatures that puddle the surface soil where NO3– accumulated after the growing season). We can see from data in Figure 14.1A and Figure 14.9B that N2O emissions were, on average, highest in January and February. The relative ratios of N2 and N2O and NO produced during these events have not yet been quantified. We estimate that about 0.04 g N⋅m–2⋅y–1 is emitted as N2 annually. Periodic measurements of NOx emissions from one site March through July 1988, were conducted by Stocker et al. (1993) using micrometeorological eddy correlation, and by Williams and Fehsenfeld (1991) using chambers. Using the information presented in Stocker et al. (1993), we estimate an annual NO emission of 0.11 g N⋅m–2⋅y–1. This estimate was made from their flux estimates of midsummer (7.2 ng N⋅m–2⋅s–1 for 30 days), summer (5.7 ng N⋅m–2⋅s–1 for 90 days), and long term (2.2 ng N⋅m–2⋅s–1 for 265 days), and summing the three periods to obtain an annual flux estimate. To improve this NO emission estimate and to observe how fluxes vary temporally as well as across the landscape, we began studies in 1994 to measure emissions throughout the year (e.g., Fig. 14.6) in a number of different sites (Martin et al., 1998). These new studies generally agree with the previous NO emission studies, in that the N lost by this mechanism represents an important part of the annual N budget of the system. The year-round measurements do suggest that fluxes are probably higher than the earlier studies indicate, with an estimated annual emission of 0.2 g NO N⋅m–2⋅y–1. This estimate comes from observations at two coarse-textured sites and two finer textured sites where fluxes averaged 7.6 ng⋅m–2⋅s–1 and 5.2 ng⋅m–2⋅s–1, respectively. Using the array of Schimel et al. (1986) of 70% coarse-textured upland soil and 30% finer textured lowland soil across the shortgrass steppe, we estimate an integrated annual emission of 0.22 g NO N⋅m–2⋅y–1. We do not, however, have an adequate assessment of interannual variation needed to compare fluxes from year to year.

Summary Our long-term, year-round measurements of CH4 and N2O fluxes show that gas fluxes vary across the shortgrass steppe landscape, and that these grasslands serve

Exchange of Trace Gases in the Shortgrass Steppe 367

as both an important sink for atmospheric CH4 and a source of N2O. Our studies also provide information for an increased understanding of the seasonal dynamics and interannual variation of CH4 uptake and N2O flux, and an appreciation of the importance of emissions of N gases (NH3, NOx, N2O) as losses from the ecosystem. These data have been used in the development and refinement of simulation models that can be used to describe fluxes of CH4, NOx,, and N2O across the shortgrass steppe in response to variations of climate, landscape, and land-use and management change (Del Grosso et al., 2000a,b; Parton et al., 2001). Important findings from the field studies include the following: 1. Emissions of NOx and N2O are important N-loss processes, and volatile N emissions likely regulate many aspects of the shortgrass steppe soil–plant system. We are lacking, however, a complete understanding of the impacts of annual NOx deposition, NH3 emissions and uptake from plants, and N losses through winter denitrification on the N budgets of the system. 2. Climate and weather are important controls over trace gas fluxes. Wintertime fluxes of both CH4 and N2O constitute a large portion of the annual flux of these gases. Because of climatic variability, there are interannual variations of the annual mean fluxes of both N2O and CH4. 3. Land use is an important control over trace gas flux. Pastures having similar land-use history and soil texture show similar seasonal and annual N2O emission and CH4 uptake rates and patterns, but there are strong differences among land-use types. Most important, conversion of grasslands to croplands promotes changes in CH4 and N2O flux. Immediately after tillage, CH4 consumption decreased by about 35% and remained at the decreased rate for at least 3 years. Nitrous oxide emissions were about eight times higher from tilled soil for about 18 months after tillage. Two to 3 years after tillage, N2O emissions averaged only 25% to 50% higher in the tilled soils than the native grassland. Long-term cropping (winter wheat) resulted in lower CH4 uptake. Recovery of cultivated soils back to perennial grasslands eventually leads to soils having CH4 oxidation and N2O emissions similar to those of native soils of the same texture and parent material. Complete recovery after tillage requires longer than 8 years but less than 50 years. 4. Under elevated CO2, none of the trace gas fluxes appear to be significantly altered over a 5-year period. However, addition of water and NH4+ NO3- to soils that had been exposed to double-ambient CO2 concentrations for five growing seasons increased ecosystem respiration and atmospheric CH4 oxidation, and decreased NO emissions. These observations suggest that methanotrophic populations were enhanced under elevated CO2 whereas soil N supply was depleted by increased plant growth (King et al., 2004; Morgan et al., 2001). 5. Soil respiration was higher in previously elevated CO2 soils after irrigation and N addition, suggesting that microbes were becoming N limited. Although midway through the 5-year experiment decomposition rates were twice as high under elevated as under ambient CO2 (Pendall et al.,

368

Ecology of the Shortgrass Steppe

2003), the N fertilization response observed here suggests that eventually microbial decomposition rates will slow, as predicted by Hu et al. (2001), leading to increased C sequestration potential. Our data confirm that fluxes of N2O and CH4 within grasslands represent an important part of the global atmospheric budget. Temperate grasslands comprise about 8% (≈11.5 × 1012 m2) of global land surface area (Bouwman, 1990). Assuming a spatial distribution in all temperate grasslands of 70% upland and 30% lowland (Schimel et al., 1986), our estimated average N2O emission and CH4 uptake rates are 1.6 μg N⋅m–2⋅hr–1 and 31.2 μg C⋅m–2⋅hr–1, or 14.2 mg N⋅m–2⋅y–1 and 314 mg C⋅m–2⋅y–1, respectively. Assuming a global temperate grassland area of 11.5 × 1012 m2, and that our grassland sites are typical of the rest of the world, N2O emissions averaged 0.16 Tg⋅y–1. This amount represents about 1.1% of annual global production (IPCC, 1992). Tropical grasslands are expected to emit N2O at several times higher annual rates (Keller et al., 1993; Matson and Vitousek, 1991; Mosier et al., 1993). Consumption of atmospheric CH4 by grassland soils averaged 3.2 Tg CH4 C⋅y–1 (4.2 Tg CH4), about 1% of CH4 production (IPCC, 1992). As Ojima et al. (1993) point out, without the aerobic soil sink for atmospheric CH4, the rate of CH4 concentration increase in the atmosphere would be about two times higher than the rate observed in the 1980s.

Acknowledgments We appreciate the invaluable assistance of Anita Kear for conducting the majority of the gas flux measurements. We also thank Brian Newkirk for his help with graphics and statistical analyses, and Becky Riggle, Chris Hegdahl, David Jensen, and Larry Tisue for their capable technical assistance. Financial support for these studies came principally from USDA–ARS and SGS LTER NSF grant DEB 0217631.

References Aguilar, R., E. F. Kelly, and R. D. Heil. 1988. Effects of cultivation on soils in northern Great Plains rangelands. Soil Science Society America Journal 52:1076–1081. Andreae, M. O., and D. S. Schimel (eds.). 1989. Exchange of trace gases between terrestrial ecosystems and the atmosphere. John Wiley, New York. Bouwman, A. F. 1990. Exchange of greenhouse gases between terrestrial ecosystems and the atmosphere, pp. 61–127. In: A. F. Bouwman (ed.), Soils and the greenhouse effect. John Wiley, New York. Bronson, K. F., and A. R. Mosier. 1994. Suppression of methane oxidation in aerobic soil by nitrogen fertilizers, nitrification inhibitors, and urease inhibitors. Biology and Fertility of Soils 17:263–268. Brumme, R., W. Borken, and S. Finke. 1999. Hierarchical control on nitrous oxide emission in forest ecosystem. Global Biogeochemical Cycles 13:1137–1148. Burke, I. C., E. T. Elliott, and C. V. Cole. 1995. Influence of macroclimate, landscape position, and management on soil organic matter in agroecosystems. Ecological Applications 5:124–131. Coffin, D. P., W. K. Lauenroth, and I. C. Burke. 1996. Recovery of vegetation in a semiarid grassland 53 years after disturbance. Ecological Applications 6:538–555.

Exchange of Trace Gases in the Shortgrass Steppe 369 Crutzen, P. J. 1983. Atmospheric interactions in homogeneous gas reactions of C, N and S containing compounds, pp. 67–112. In: B. Bolin and R. B. Cook (eds.), The major biogeochemical cycles and their interactions. SCOPE vol. 21. John Wiley, New York. Davidson, E. A., and J. L Hackler. 1994. Soil heterogeneity can mask the effects of ammonium availability on nitrification. Soil Biology Biochemistry 26:1449–1453. Davidson, E. A., P. A. Matson, P. M. Vitousek, R. Riley, K. D. Dunkin, G. Garcia-Mendez, and J. M. Maass. 1993. Processes regulating soil emissions of NO and N2O in a seasonally dry tropical forest. Ecology 74:130–139. Davidson, E. A., J. M. Stark, and M. K. Firestone. 1990. Microbial production and consumption of nitrate in an annual grassland. Ecology 71:1968-1975. Del Grosso, S. J., W. J. Parton, A. R. Mosier, D. S. Ojima, A. E. Kulmala, and S. Phongpan. 2000a. General model for N2O and N2 gas emissions from soils due to denitrification. Global Biogeochemical Cycles 14:1045–1060. Del Grosso, S .J., W. J. Parton, A. R. Mosier, D. S. Ojima, C. S. Potter, W. Borken, R. Brumme, K. Butterbach-Bahl, P. M. Crill, K. Dobbie, and K. A. Smith. 2000b. General CH4 oxidation model and comparisons of CH4 oxidation in natural and managed systems. Global Biogeochemical Cycles 14:999–1019. Elliott, E. T., and C. V. Cole. 1989. A perspective on agroecosystem science. Ecology 70:1597–1602. Epstein, H. E., I. C. Burke, A. R. Mosier, and G. L. Hutchinson. 1998. Plant functional type effects on trace gas fluxes in the shortgrass steppe. Biogeochemistry 42:145–168. Firestone, M. K., and E. A. Davidson. 1989. Microbiological basis of NO and N2O production and consumption in soil, pp. 7–21. In: M. O. Andreae and D. S. Schimel (eds.), Exchange of trace gases between terrestrial ecosystems and the atmosphere. John Wiley, New York. Hu, S., F. S. Chapin, III, M. K. Firestone, C. B. Field, and N. R. Chiariello. 2001. Nitrogen limitation of microbial decomposition in a grassland under elevated CO2. Nature 409:188–191. Hungate, B. A., F. S. Chapin, H. Zhong, E. A. Holland, and C. B. Field. 1997a. Stimulation of grassland nitrogen cycling under carbon dioxide enrichment. Oecologia 109:149–153. Hungate, B. A., E. A. Holland, R. B. Jackson, F .S. Chapin, H. A. Mooney, and C. B. Fields. 1997b. The fate of carbon in grasslands under carbon dioxide enrichment. Nature 388:576–579. Hungate, B. A., C. P. Lund, H. L. Pearson, and F. S. Chapin. 1997c. Elevated CO2 and nutrient addition alter soil N cycling and N trace gas fluxes with early season wet-up in a California annual grassland. Biogeochemistry 37:89–109. Hutchinson, G. L., and A. R. Mosier. 1981. Improved soil cover method for field measurement of nitrous oxide fluxes. Soil Science Society of America Journal 45:311–316. Ihori, T., I. C. Burke, W. K. Lauenroth, and D. P. Coffin. 1995. Effects of cultivation and abandonment on soil organic matter in northeastern Colorado. Soil Science Society of America Journal 59:1112–1119. Ineson, P., P. A. Coward, and U. A. Hartwig. 1998. Soil gas fluxes of N2O, CH4 and CO2 beneath Lolium perenne under elevated CO2: The Swiss free air carbon dioxide enrichment experiment. Plant and Soil 198:89–95. IPCC. 1992. Climate change 1992. The supplementary report to the IPCC Scientific Assessment, Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, UK. IPCC. 1996. Climate change 1995. The science of climate change. Intergovernmental panel on climate change. Cambridge University Press, Cambridge, UK.

370

Ecology of the Shortgrass Steppe

IPCC. 2007. Intergovernmental panel on climate change. Climate change: The physical basis. Summary for Policy makers. Contribution of Working Group I to the fourth assessment report of the IPCC. Online. Available at www.ipcc.ch. Johansson, C., and E. Sanhueza. 1988. Emission of NO from savanna soils during rainy season. Journal Geophysical Research 93:14193–14198. Keller, M., E. Veldkamp, A. M. Weltz, and W. A. Reiners. 1993. Effect of pasture age on soil trace gas emissions from a deforested area of Costa Rica. Nature 365:244–246. King, J. Y., A. R. Mosier, J. A. Morgan, D. R. LeCain, D. G. Milchunas, and W. J. Parton. 2004. Plant nitrogen dynamics in shortgrass steppe under elevated atmospheric carbon dioxide. Ecosystems 7:147–160. Langford, A. O., F. C. Fehsenfeld, J. Zachariassen, and D. S. Schimel. 1992. Gaseous ammonia fluxes and background concentrations in terrestrial ecosystems of the United States. Global Biogeochemical Cycles 6:459–483. LeMer, J., and P. Roger. 2001. Production, oxidation, emission and consumption of methane by soils: A review. European Journal of Soil Biology 37:25–50. Li, C., S. Frolking, and T. A. Frolking. 1992. A model of nitrous oxide evolution from soil driven by rainfall events, 1: Model structure and sensitivity. Journal Geophysical Research 97D:9759–9776. Martin, R. E., M. C. Scholes, A. R. Mosier, D. S. Ojima, E. A. Holland, and W. J. Parton. 1998. Controls on annual emissions of nitric oxide from soils of the Colorado shortgrass steppe. Global Biogeochemical Cycles 12:81–91. Matson, P. A., and P. M. Vitousek. 1990. Ecosystem approach to a global nitrous oxide budget. BioScience 40:667–672. McLain, J. E. T., T. B. Kepler, and D. M. Ahmann. 2002. Belowground factors mediating changes in methane consumption in a forest soil under elevated CO2. Global Biogeochemical Cycles 16(23):1–14. Milchunas, D. G., W. J. Parton, D. S. Bigelow, and D. S. Schimel. 1988. Factors influencing ammonia volatilization from urea in soils of the shortgrass steppe. Journal Atmospheric Chemistry 6:323–340. Monson, R. K., and E. A. Holland. 2001. Biospheric trace gas fluxes and their control over tropospheric chemistry. Annual Review of Ecology and Systematics 32:547–576. Morgan, J. A., D. R. LeCain, A. R. Mosier, and D. G. Milchunas. 2001. Elevated CO2 enhances water relations and productivity and affects gas exchange in C3 and C4 grasses of the Colorado shortgrass steppe. Global Change Biology 7:451–466. Morgan, J. A., A. R. Mosier, D. G. Milchunas, D. R. LeCain, J. A. Nelson, and W. J. Parton. 2004. CO2 enhances productivity, alters species composition, and reduces digestibility of shortgrass steppe vegetation. Ecological Applications 14:208–219. Mosier, A. R., G. L. Hutchinson, B. R. Sabey, and J. Baxter. 1982. Nitrous oxide emissions from barley plots treated with ammonium nitrate or sewage sludge. Journal of Environmental Quality 11:78–81. Mosier, A. R., J. A. Morgan, J. Y. King, D. LeCain, and D. G. Milchunas. 2002a. Soil atmosphere exchange of CH4, CO2, NOx and N2O in the Colorado shortgrass steppe under elevated CO2. Plant and Soil 240:200–211. Mosier, A. R., J. A. Morgan, J. Y. King, D. LeCain, and D. G. Milchunas. 2002b. Soil atmosphere exchange of CH4, CO2, NOx, and N2O in the Colorado shortgrass steppe under elevated CO2. Presented at the 7th Scientific Conference of the International Committee on Global Atmospheric Chemistry. Herakliion, Crete, Greece, September 18–25. Mosier, A. R., and W. J. Parton. 1985. Denitrification in a shortgrass prairie: A modeling approach, pp. 441–451. In: D. E. Caldwell, J. A. Brierley, and C. L. Brierley (eds.), Planetary Ecology. Van Nostrand Reinhold, New York.

Exchange of Trace Gases in the Shortgrass Steppe 371 Mosier, A. R., W. J. Parton, and S. Phongpan. 1998. Long term large N and immediate small N addition effects on trace gas fluxes in the Colorado shortgrass steppe. Biology & Fertility of Soils 28:44–50. Mosier, A. R., W. J. Parton, D. W. Valentine, D. S. Ojima, D. S. Schimel, and J. A. Delgado. 1996. CH4 and N2O fluxes in the Colorado shortgrass steppe: I. Impact of landscape and nitrogen addition. Global Biogeochemical Cycles 10:387–399. Mosier, A. R., W. J. Parton, D. W. Valentine, D. S. Ojima, D .S. Schimel, and O. Heinemeyer. 1997. CH4 and N2O fluxes in the Colorado shortgrass steppe: 2. Long term impact of land use change. Global Biogeochemical Cycles 11:29–42. Mosier, A. R., E. Pendall, and J. A. Morgan. 2003. Soil–atmosphere exchange of CH4, CO2, NOx, and N2O in the Colorado shortgrass steppe following five years of elevated CO2 and N fertilization. Atmospheric Chemistry and Physics Discussions 3:2691–2706. Mosier, A. R., and D. S. Schimel. 1991. Influence of agricultural nitrogen on atmospheric methane and nitrous oxide. Chemistry & Industry 23:874–877. Mosier, A. R., D. S. Schimel, D. W. Valentine, K. F. Bronson, and W. J. Parton. 1991. Methane and nitrous oxide fluxes in native, fertilized, and cultivated grasslands. Nature 350:330–332. Mosier, A., D. W. Valentine, D. S. Schimel, W. J. Parton, and D. S. Ojima. 1993. Methane consumption in the Colorado short grass steppe. Mitteilungen der Deutschen Bodenkundlichen Gesellschaft 69:219–226. Ojima, D., A. Mosier, S. Del Grosso, and W. J. Parton. 2000. TRAGNET analysis and synthesis of trace gas fluxes. Global Biogeochemical Cycles 14:995–997. Owensby, C. E., P. I. Coyne, and L. M. Auen. 1993. Nitrogen and phosphorus dynamics of a tall grass prairie ecosystem exposed to elevated carbon dioxide. Plant Cell Environment 16:843–850. Parton, W. J., E. A. Holland, S. J. Del Grosso, M. D. Hartman, R. E. Martin, A. R. Mosier, D. S. Ojima, and D. S. Schimel. 2001. Generalized model for NOx and N2O emissions from soils. Journal of Geophysical Research 106(D15):17403–17420. Parton, W. J., A. R. Mosier, D. S. Ojima, D. W. Valentine, D. S. Schimel, K. Weier, and A. E. Kulmala. 1996. Generalized model for N2 and N2O production from nitrification and denitrification. Global Biogeochemical Cycles 10:401–412. Parton, W. J., A. R. Mosier, and D. S. Schimel. 1988. Rates and pathways of nitrous oxide production in a shortgrass steppe. Biogeochemistry 6:45–48. Pendall, E., S. Del Grosso, J. Y. King, D. R. LeCain, D. G. Milchunas, J. A. Morgan, A. R. Mosier, D. S. Ojima, W. A. Parton, P. P. Tans, and J. W. C. White. 2003. Elevated atmospheric CO2 effects and soil water feedbacks on soil respiration components in a Colorado grassland. Global Biogeochemical Cycles 17:1–13. Pendall, E., A. R. Mosier, and J. A. Morgan. 2004. Rhizodeposition stimulated by elevated CO2 in a semiarid grassland. New Phytologist 162:447–458. Reeburgh, W. S., S. C. Whalen, and J. J. Alpern. 1993. The role of methylotrophy in the global methane budget, pp. 1–14. In: J. C. Murrell and D. P. Kelly (eds.), Microbial growth on C1 compounds. Proceedings of the 7th International Symposium. Intercept, Andover, UK. Russel, J. C. 1929. Organic matter problems under dry-farming conditions. Agronomy Journal 21:960–969. Schimel, D. S., W. J. Parton, F. J. Adamsen, R. G. Woodmansee, R. L. Senft, and M. A. Stillwell. 1986. The role of cattle in the volatile loss of nitrogen from a shortgrass steppe. Biogeochemistry 2:39–52.

372

Ecology of the Shortgrass Steppe

Schlesinger, W. H., and A. E. Hartley. 1992. A global budget for atmospheric NH3. Biogeochemistry 15:191–211. Smith, K. A., T. Ball, F. Conen, K. E. Dobbie, J. Massheder, and A. Rey. 2003. Exchange of greenhouse gases between soil and atmosphere: Interactions of soil physical factors and biological processes. European Journal of Soil Science 54:779–791. Steudler, P. A., R. D. Bowden, J. M. Melillo, and J. D. Aber. 1989. Influence of nitrogen fertilization on methane uptake in temperate forest soils. Nature 341:314–331. Stillwell, M. A. 1983. The effect of bovine urine on the nitrogen cycle in a shortgrass prairie. PhD diss., Colorado State University, Fort Collins, Colo. Stocker, D. W., D. H. Stedman, K. F. Zeller, W. J. Massman, and D. G. Fox. 1993. Fluxes of nitrogen oxides and ozone measured by eddy correlation over a shortgrass prairie. Journal Geophysical Research 98:12619–12630. Tieszen, H., J. W. B. Stewart, and J. R. Bettany. 1982. Cultivation effects on the amounts and concentration of carbon, nitrogen, and phosphorus in grassland soils. Agronomy Journal 74:831–835. Tortoso, A. C., and G. L. Hutchinson. 1990. Contributions of autotrophic and heterotrophic nitrifiers to soil NO and N2O emissions. Applied Environmental Microbiology 56:1799–1805. Valentine, D. W., A. R. Mosier, R. W. Lober, and J. W. Doran. 1993. Methane uptake in grassland soils: Concentration profiles, flux measurements, and modeling. Bulletin Ecological Society of America 74:466. Williams, E. J., and F. C. Fehsenfeld. 1991. Measurement of soil nitrogen oxide emissions at three North American ecosystems. Journal Geophysical Research 96:1033–1042. Williams, E. J., G. L. Hutchinson, and F. C. Fehsenfeld. 1992. NOx and N2O emissions from soil. Global Biogeochemical Cycles 6:359–388. Zachariassen, J., and D. S. Schimel. 1991. Ammonia exchange above grassland canopies. EOS transactions. American Geophysical Union 72:110.

15 The Shortgrass Steppe and Ecosystem Modeling William J. Parton Stephen J. Del Grosso Ingrid C. Burke Dennis S. Ojima

E

cological modeling has played a key role in scientific investigations of the SGS LTER during the past several decades. The SGS LTER site, focused initially on the Central Plains Experimental Range (CPER), was the main grassland research site for the Grassland Biome component of the U.S. IBP effort (Lauenroth et al., this volume, chapter 1). Initial development of ecosystem models occurred from 1970 to 1975 as part of the IBP. All the U.S. IBP projects (grassland, tundra, desert, deciduous forest, and coniferous forest biomes) included research on the development of ecosystem models, with the goals of using models to help formulate and interpret field experiments, and of projecting the impact of changes in management practices on ecosystem dynamics. Models were developed as part of the Grassland Biome project (Bledsoe et al., 1971; Innis, 1978), and included modeling specialists who worked with research biologists on the development and formulation of the ecosystem models. The modeling activities of the U.S. IBP Grassland Biome project included developing the ELM Grassland model (Innis, 1978). The ELM model was a complex process-oriented model that was intended to be used at all the Grassland Biome sites in the United States. This model was developed by postdoctoral fellows who were to formulate the different submodels, and then link the submodels using software that was developed as part of the program. The submodels included a plant production submodel, a cattle production submodel, a linked nutrient cycling and soil organic matter submodel, a grasshopper dynamics submodel, and a soil temperature and water submodel. Biophysical and biological data from the different sites were collected to develop and test the model. Model development was constrained by lack of knowledge about the biological processes

373

374

Ecology of the Shortgrass Steppe

that control ecosystem behavior, and by lack of appropriate data to test the ability of the model to simulate ecosystem responses to changes in grazing and fertility management practices. However, the ELM Grassland model was quite successful at investigating the interactions of different components of the ecosystem, and at helping to formulate new research efforts. Unfortunately, the model was very difficult to use because of the amount of time required to make computer runs, and the large amount of data required to verify the response of the model to changes in the environment. Ultimately, the knowledge developed as part of building the ELM Grassland model was successfully used to develop the next generation of grassland ecosystem models. The SGS LTER program and its IBP Grassland Biome predecessor have had a major impact on the development of grassland ecosystem models since the ELM model was formulated in the 1970s. Important models that have been developed during the last 30 years using data from the SGS LTER site include the ROOT, PHOENIX, CENTURY, Foodweb, Grassland Ecosystem Model (GEM), SPUR, STEPPE, and DAYCENT models. Here, we will briefly review these models. The ROOT model (Parton et al., 1978) used detailed root growth data (root window data) from the SGS LTER site to formulate a mechanistic root growth model that simulated the response of roots to changes in soil temperature, water, and fertility. The PHOENIX model (McGill et al., 1981) simulated soil nutrient cycling and carbon (C) dynamics for grasslands and was developed using many of the concepts included in the ELM nutrient cycling submodel. The CENTURY model (Parton et al., 1987) is a generalized grassland ecosystem model that was developed to simulate plant production, nutrient cycling, and soil organic matter dynamics for grasslands and agroecosystems in the Great Plains. The CENTURY model used experience gained from the development of the PHOENIX, ROOT, and ELM models to develop a simplified ecosystem model that could simulate grassland and crop ecosystem dynamics at the regional scale for the Great Plains. Hunt et al. (1984) developed the Foodweb model to simulate nutrient cycling, C dynamics, and soil animal dynamics using extensive data sets collected at the SGS LTER site during the 1980s. The gap dynamics model, STEPPE (Coffin and Lauenroth, 1990), was developed to represent shortgrass steppe plant community dynamics; this model is described at length in Peters and Lauenroth (chapter 7, this volume). Hunt et al. (1991) developed the GEM, and Hanson et al. (1988) developed the SPUR grassland ecosystem model. Both of these models were developed using data sets from the SGS LTER, in addition to experimental data, and were designed to simulate the impact of climatic change and increases in atmospheric CO2 on grassland ecosystem dynamics. Most recently, detailed process-oriented nutrient cycling, trace gas flux, and soil organic matter data collected during the 1990s have been used to develop and test the DAYCENT model. DAYCENT has been used to simulate trace gas fluxes and ecosystem dynamics for grasslands and agroecosystems in the Great Plains and the Midwest, and to estimate nitrous oxide (N2O) emissions from agricultural soils for the U.S. National Greenhouse Gas Inventory (Del Grosso et al., 2001, 2006; Kelly et al., 2000; Parton et al., 1998). The detailed process-oriented plant production, nutrient cycling, soil organic matter dynamics, and trace gas flux data collected at the SGS LTER site have played

The Shortgrass Steppe and Ecosystem Modeling 375

a critical role in the development of most of the grassland ecosystem models used in the United States during the past 30 years. Ecosystem models developed in conjunction with the SGS LTER site have been used extensively to evaluate the impacts of changes in grassland and cropping management practices, climate, and atmospheric CO2 level. The models have been used for impact assessment at the site-specific level as well as at regional and global scales. Ecosystem models (e.g., CENTURY and SPUR) developed at the SGS LTER site were among the first models used to extrapolate climate change impacts on C cycling at regional and global scales (Parton et al., 1994, 1995; Schimel et al., 1991). Model results from Pepper et al. (2005) suggested that increasing atmospheric CO2 levels and air temperature will increase plant growth in shortgrass steppe, tallgrass prairie, and boreal forest ecosystems. However, decomposition rates also increased and net C uptake was close to neutral unless nitrogen (N) inputs were increased, in which case all three systems became net sinks of C for the duration of the simulations (2000–2100). Likewise, more recent model simulations from Parton et al. (2007) predict that increasing atmospheric CO2 levels and air temperature will increase plant growth in semiarid grasslands in southeast Wyoming, particularly during wet years. In agreement with previous model results, increasing atmospheric CO2 levels and air temperature also increased decomposition rates, and hence soil organic matter declined slightly during the period simulated (2000–2015). In this chapter we will first present a short review of how models developed for the shortgrass steppe have been used for testing our understanding of grassland ecosystem processes, and for impact assessment during the past 30 years. Here, our goal is to illustrate how the models have been used to advance our understanding of grasslands dynamics. Second, we will apply the DAYCENT model to assess the vulnerabilities of the shortgrass steppe to potential climate change during the next 100 years. This analysis will consider the impacts of changing precipitation, temperature, atmospheric CO2, and N deposition levels.

Site-Specific Impact Assessment Models developed originally for the shortgrass steppe have been used for numerous environmental impact studies. Grassland ecosystem models have been used to evaluate the environmental impacts of insect outbreaks, range management schemes, and climatic patterns on soil organic matter dynamics, grassland production, nutrient cycling, and secondary production. Some of the earliest work included the use of the ELM model (Innis, 1978) to simulate grassland ecosystem responses to potential climate change associated with both cloud seeding for hail suppression and the deployment of a fleet of supersonic transports (Parton and Smith, 1974a, b). The ELM model was also used to assess the ecosystem impact of overgrazing in the Sahel region of Africa during the drought of the 1970s (Parton and Schnell, 1975). The ELM model and other grassland ecosystem models have been used to evaluate grazing and fire management implications for an Oklahoma tallgrass prairie (Holland et al., 1992; Ojima et al., 1990; Parton and

376

Ecology of the Shortgrass Steppe

Risser, 1980; Parton et al., 1980), to develop an impact assessment of management practices used for strip-mine reclamation (Parton et al., 1979), and to evaluate the impact of grasshopper and caterpillar grazing on plant production and nutrient cycling (Capinera et al., 1983a, b). Numerous studies have applied CENTURY to scenarios of cultivation management to evaluate ecosystem responses at a site level, or across multiple sites in the Great Plains (e.g., Campbell et al., 2005). During the early 1990s, there was substantial interest in evaluating the impact of potential climatic change associated with increasing atmospheric CO2. The GEM model (Hunt et al., 1991), CENTURY model (Schimel et al., 1990), and SPUR model (Hanson et al., 1988) were used to simulate the impact of increased atmospheric CO2-induced potential climate changes on plant production, nutrient cycling, and animal production for sites in the Great Plains. Results indicated that plant production was most sensitive to changes in precipitation, whereas soil organic matter dynamics, decomposition, and nutrient cycling were sensitive to changes in both precipitation and temperature (increasing temperature and precipitation increased both nutrient cycling and decomposition of litter and soil organic matter). The GEM model results (Hunt et al., 1991) suggested that increasing atmospheric CO2 levels would increase plant production and soil organic matter levels, and reduce the negative impact of potential decreases in precipitation.

Regional Impact Assessment The CENTURY model was one of the first ecosystem models to be applied at a regional scale (Parton et al., 1987) or to be linked to spatially explicit input data organized in a geographic information system to simulate regional ecosystem dynamics (Burke et al., 1990). This structure takes information from spatial and temporal data of soils, climate, and land use to simulate regional responses of ecosystems to changes in management and environmental variables. Burke et al. (1991) and Schimel et al. (1991) used the CENTURY model to simulate ecosystem dynamics at the regional scale with a primary focus on the impact of potential climate changes on ecosystem dynamics within the Great Plains. The results from these studies suggested that increasing temperature may result in substantial loss of soil C, and that plant production was positively correlated to changes in precipitation. The paper by Schimel et al. (1991) also showed that the simulated regional patterns in grassland production were well correlated with the observed regional patterns in the NDVI (remotely sensed index of plant production). Burke et al. (1997) used CENTURY-simulated regional patterns of N mineralization and other ecosystem parameters to evaluate the relative importance of the factors that control plant production at the regional scale. Burke et al. (1991) have found that the regional C consequences of landuse management may be greater than those of climate change. Parton et al. (2005) recently used a regional application of CENTURY to simulate the large-scale responses of grasslands to cultivation management. Their results showed large-scale losses of soil C with early cultivation practices. Plowing, which enhances decomposition rates, and low crop residue inputs associated

The Shortgrass Steppe and Ecosystem Modeling 377

with dryland wheat-fallow cropping both contributed to soil C losses. However, Great Plains soils converted from dryland to irrigated cropping since the 1970s have become a C sink—estimated at 21.3 Tg C for the region. Del Grosso et al. (2006) applied DAYCENT to cropped soils to evaluate national-scale consequences of cropping on NO2 production. Last, CENTURY has been incorporated into economic models to estimate the regional-scale economic potentials for agricultural C sequestration (Antle et al., 2007).

Land–Atmosphere Interactions Our understanding of the interactions between the atmosphere and the land surface (referring to the soil, vegetation, water system) is critical to estimating the vulnerability of key natural resources to climate and land-use changes (Pielke et al., 1997). Changes in these interactions affect mesoscale physical and chemical climate, water basin hydrology, and ecological properties, such as vegetation composition, disturbance regime, and biogeochemical cycles. Terrestrial biospheric processes respond strongly to atmospheric temperature, humidity, CO2 levels, N deposition, precipitation, and radiative transfers. Grassland simulation models have been important for understanding the biophysical feedbacks that couple the land surface to the atmosphere through processes controlling energy and water exchanges (Eastman et al., 1998; Vidale et al., 1997; Walko et al., 2000). These feedbacks operate rapidly and are estimated many times each hour. Biogeochemical and ecosystem interactions with atmospheric processes have been implemented using the CENTURY-RAMS model (Lu et al., 2001; Pielke et al., 1997). The coupled CENTURY biogeochemical and RAMS atmospheric model showed that seasonal vegetation growth patterns strongly influence surface water and energy exchanges, and regional climate patterns. These research efforts have been directed toward developing a better understanding of how biophysical coupling to the atmosphere changes over time, as the land surface and ecosystem processes change the constraints on water, C, and energy fluxes, and on hydrological and ecological processes (Ojima et al., 1991; Schimel et al., 1990, 1994).

Long-Term Ecosystem Response to Environmental Change: A Vulnerability Assessment In this section we use a model developed from long-term data at the SGS LTER to assess the vulnerability of the shortgrass steppe ecosystem to possible future long-term changes in the environment. Our analysis focuses on CO2 increases, associated climate change, and increased N deposition. Atmospheric concentration of CO2 has increased from a preindustrial level of ≈280 ppm to ≈380 ppm at present, and is expected to exceed ≈650 ppm by 2100 unless anthropogenic CO2 emissions fall below 1990 levels by the end of this century (IPCC, 2001, 2007). Current predictions are that temperatures will increase between 1.8 and 4.0 ºC (IPCC, 2007). Increased atmospheric CO2 concentration

378

Ecology of the Shortgrass Steppe

has direct and indirect effects on many important ecosystem parameters. In addition to climate change resulting from increased radiative forcing from elevated greenhouse gases, CO2 levels also influence plant growth. Three mechanisms are responsible for the CO2 fertilization effect observed in both laboratory and field experiments (Körner, 2006; Woodward, 2002). Photosynthesis rates can increase, stomatal conductance can decrease, and C-to-nutrient ratios may widen (Gifford et al., 2000; Morgan et al., 2004b). Climate change related to increased greenhouse gases in the atmosphere has the potential to alter net primary production (NPP) by changing water and temperature stress. Climate change also can alter soil C levels and N mineralization rates (Hu et al., 2001), and thus influence NPP indirectly. Open-top chamber experiments performed at the shortgrass steppe show significant effects of elevated CO2 on NPP and soil water content (Morgan et al., 2001). Net primary production increased by 25% to 50% under elevated CO2 (720 ppm), and soil water content was higher during the 4 years of the experiment. The primary mechanism was likely increased water use efficiency, with minor increases in photosynthesis rates and in C-to-N ratios for some species. Morgan et al. (2001) conclude that drier systems will likely show higher NPP increases under elevated CO2 than less water-stressed systems. With elevated CO2 levels, leaf stomata may close to some extent, but will still maintain some CO2 diffusion, thus resulting in higher photosynthetic rates than in low CO2 conditions. However, the long-term effects of increased CO2 on plant growth rates, N availability, and soil organic matter levels are uncertain. Climate data show a trend of increasing temperature since approximately 1950 for the Great Plains region, particularly nighttime minimum temperatures (Alward et al., 1999; Pielke et al., 2000). Rainfall was variable during the 20th century, and future changes in rainfall patterns are uncertain. Model simulations are required to quantify the sensitivity of ecosystem responses to these uncertain drivers and to simulate longterm changes that cannot be addressed using field experiments.

DAYCENT and CENTURY Model Descriptions DAYCENT (Del Grosso et al., 2001; Kelly et al., 2000; Parton et al., 1998) is the daily time step version of the CENTURY model (Parton et al., 1994). DAYCENT and CENTURY simulate exchanges of C, nutrients, and trace gases among the atmosphere, soil, and vegetation (Fig. 15.1). Both models simulate decomposition and nutrient mineralization of plant litter and soil organic matter, plant growth and senescence, and soil water and temperature fluxes. CENTURY has a monthly time step, and uses monthly maximum/minimum temperature and precipitation data as input. Site-specific data such as soil texture and hydraulic properties (one value for all soil layers), and vegetation type are also required as input data. DAYCENT requires daily maximum/minimum temperature and precipitation data, and soil texture, bulk density, and hydraulic properties must be specified for each soil layer. Both models simulate three pools of soil organic matter (active, slow, passive) that have different maximum turnover rates and two mineral N

The Shortgrass Steppe and Ecosystem Modeling 379

DAYCENT Model

N Gas 0-1 cm 1-4 cm 4-15 cm 15-30 cm etc.

H2Osoil Tsoil

St

om

Plant Components Leaves Fine Roots Branches Large Wood Large Roots

0-1 cm 1-4 cm 4-15 cm 15-30 cm etc.

_

NO3

ts N inpu L V , PPT ,

NH4+

0-1 cm 0-15 1-4 cm cm 4-15 cm 15-30 cm etc.

S C:N

SOM Active

C:N

CO 2

S

0.5-1 yr

Slow

10-50 yr

S

p

Passive

1000-5000 yr

com

Dead Plant Decomp Rh S Material Structural Metabolic CO 2

Rh

De

h Deat

C:N

ET

CO 2

N min

N uptake

NPP

Den

V

Nit

S, Rh

= C, N flows = feedbacks; information flows = control process Processes designated by italics: H2Osoil = soil water content Stom = stomatal conductance Tsoil = soil temperature Death = plant component death Decomp = decomposition S = soil texture C:N = carbon:nitrogen ratio of material N inputs = N fixation, deposition, fertilization Nit = nitrification V = vegetation type Den = denitrification SOM = soil organic matter N min = N mineralization L = land use ET = evapotranspiration R h = heterotrophic respiration N Gas = N2O, NOx , N2

Figure 15.1 Flow diagram for the DAYCENT biogeochemical model. (From Parton et al. [2007].)

pools (ammonium and nitrate). Ammonium (NH4+) is assumed to be confined to the top 15 cm of soil whereas nitrate (NO3-) is mobile and can leach out the bottom of the profile. In addition to having finer temporal resolution, soil layers are more refined in DAYCENT. Typically, soil layers are 15 or 30 cm thick in CENTURY but are 15 cm or less (for surface layers) in DAYCENT. In both models, users can change the soil layer structure and the depth of the profile. Typical management and disturbance events (plowing, burning, harvesting, fertilizing, irrigating, and so forth) can be readily implemented in both models. In terms of processes, the major difference between the models is that DAYCENT explicitly represents the processes (nitrification and denitrification) that lead to N2O, NOx, and N2 emissions, whereas CENTURY assumes that a constant proportion of available N in each time step is lost as N gas without distinguishing between the different N gas

380

Ecology of the Shortgrass Steppe

species. Unlike CENTURY, DAYCENT includes a submodel to represent uptake of atmospheric methane (CH4) by soil microbes. The soil water and temperature submodels are also much more detailed in DAYCENT. The primary motivation for developing DAYCENT was reliable simulation of N gas emissions from soils. The response of N gas emissions to soil water content and texture is nonlinear, and N gas emissions are highly pulse driven. It is not possible to correctly simulate the short-term (daily timescale) conditions that result in pulses of N gas flux using monthly scale driving variables. The shorter time step allows DAYCENT to include submodels for nitrification and denitrification to simulate N gas emissions. Although denitrification is generally not an important part of the annual N budget in the shortgrass steppe, there are events that may lead to significant N losses through denitrification, particularly in high clay content soils (Mosier et al., 1996; Parton et al., 1988). Del Grosso et al. (2000) developed the DAYCENT denitrification submodel for nitrous oxide (N2O) and NOx gas emissions from soils from observations of N gas losses from incubations of intact and disturbed soil cores. Nitrous oxide emissions from denitrification are a function of soil NO3– concentration, water-filled pore space (WFPS), heterotrophic respiration, and texture. The model assumes that denitrification rates are controlled by the availability of soil NO3–(electron acceptor), labile C compounds (electron donor), and oxygen (competing electron acceptor). Heterotrophic soil respiration is used as a proxy for labile C availability, whereas oxygen availability is a function of soil physical properties that influence gas diffusivity, soil WFPS, and oxygen demand. Oxygen demand (indicated by respiration rates) varies inversely with a soil gas diffusivity coefficient, which is regulated by soil porosity and pore size distribution. Model inputs include soil heterotrophic respiration rate, texture, NO3– concentration, and WFPS. The model selects the minimum of the NO3– and CO2 functions to establish a maximum potential denitrification rate for particular levels of electron acceptor and C substrate, and accounts for limitations of oxygen availability, to estimate daily N2 + N2O flux rates. The output of the ratio function is combined with the estimate of total N gas flux rate to infer N2O emission. DAYCENT Simulations and Results We used DAYCENT to investigate the effects of changes in precipitation, temperature, CO2 concentration, and N deposition on some key ecosystem properties and processes. Soil and climate data from the CPER were used to drive DAYCENT for these simulations. We simulated a sandy loam soil and used long-term (1948– 2000) weather data to initialize the C and nutrient pools in the model. To validate the parameters in the model used to simulate the CO2 effect, we compared simulated and observed NPP data for ambient and elevated (720 ppm) CO2. DAYCENT showed an approximate 30% increase in NPP for the shortgrass steppe, which agrees with data collected from open-top chamber elevated CO2 experiments performed at the CPER (Morgan et al., 2001). To establish a control to compare with altered CO2, climate, and N deposition, we simulated ambient CO2 and nonmodified climate from 2001 until 2100 by

The Shortgrass Steppe and Ecosystem Modeling 381

recycling 10 years (1991–2000) of actual weather data. We then simulated five alternative scenarios (see Table 15.1 for input variables changed) to represent the effects of elevated CO2, increased temperature, increased precipitation, decreased precipitation, and increased N deposition. We included moderate grazing during the growing season (April–October) in all scenarios. The input variables were linearly ramped up (or down) in 10-year increments. For example, with the water treatment, precipitation was increased 2% above the control during 2001 to 2010, 4% from 2011 to 2020, and 20% from 2091 to 2100. We did not alter the number of precipitation events, but increased or decreased the size of the existing events by the appropriate percentage. To represent the effect of increased CO2, transpiration was assumed to decrease and the C-to-N ratio of aboveground biomass was assumed to increase. These effects are based on data from the shortgrass steppe showing that soil water contents were higher and aboveground biomass N concentrations tended to be lower in plots from elevated CO2 chambers compared with plots from ambient CO2 chambers (Morgan et al., 2001). Increased air temperature was simulated, with minimum temperature being increased to a greater extent than maximum temperature, because data show that temperatures have been increasing in the Great Plains, and that nighttime minimum temperatures have increased to a greater extent than daily maximum temperatures (Alward et al., 1999; Pielke et al., 2000). We summarized critical ecosystem output variables, including annual and 10-year average values for aboveground net primary production (ANPP), belowground net primary production (BNPP), system C (defined as the sum of C in soil organic matter, surface organic matter, litter, and live biomass), N2O emissions, NOx emissions, and CH4 uptake for each scenario (Table 15.1). Increased N, elevated CO2, and increased precipitation all increased total NPP and ANPP (Fig. 15.2A, B). Decreased precipitation led to decreased NPP, and increased temperature had little effect on NPP. Elevated CO2, increased temperature, and increased precipitation each led to slightly lower system C compared with the control (Fig. 15.2C). Increased N deposition led to large increases in system C, whereas decreased precipitation led to a small increase in system C. Elevated

Table 15.1 Control and Alternative Scenarios Simulated for the Shortgrass Steppe Scenario Control CO2 ⫹T ⫹H2O ⫺H2O ⫹N

Transpiration, % — p 30 — — — —

Aboveground C:Na, %

Tmin, ºC

Tmax, ºC

PPT, %

N, g·m–2

— n 10 — — —

— — n2 — —

— — n1 — —

— — — n 20 p 20

— — — — —

—

—

—

—

1

a carbon-to-nitrogen ratio for aboveground biomass. Each row represents a different simulation and the columns show how the model inputs were changed relative to the control. The inputs were linearly ramped up (or down) from 2001 to 2100. PPT, precipitation; T, temperature.

382

Ecology of the Shortgrass Steppe

TNPP (g C m-2 y-1)

150

(A)

125 100

control CO2 +T + H 2O

75

- H 2O +N

50

ANPP (g C m-2 y-1)

80

60 50 40 30 4000

System C (g C m-2)

(B)

70

3900 3800 3700

(C)

3600 3500 3400 3300 3200 3100 2000

2020

2040

2060

2080

2100

Year Figure 15.2 Time series of a 10-year average total productivity, aboveground productivity, and system C for shortgrass steppe simulations under elevated CO2, climate change, and increased N deposition. System C is the sum of C in SOC, surface organic matter, litter, and live biomass. TNPP, total NPP.

CO2 and increased precipitation both substantially increased NPP (Fig. 15.3A), but temperature had little impact. Decreased precipitation resulted in decreased NPP. Increased N deposition resulted in the largest ANPP enhancement and also increased BNPP. Increased N, elevated CO2, and increased precipitation each led to higher N2O and NOx emissions (Fig. 15.3B). Increased temperature was found to increase NOx emissions to a much greater extent than N2O emissions. Decreased precipitation decreased both N2O and NOx emissions substantially. Methane uptake rates were not strongly affected by any of the treatments except for N addition, which decreased CH4 uptake.

The Shortgrass Steppe and Ecosystem Modeling 383 70

(A) 10 year average (2091-2100) change relative to control

60

% delta control

50 ANPP BNPP

40 30 20 10 0 -10 -20 70 60

(B)

% delta control

50 N 2O NOx CH4

40 30 20 10 0 -10 -20

CO2

+T

+ H2O

- H2O

+N

Figure 15.3 Effects of DAYCENT simulations for the shortgrass steppe on NPP and trace gas fluxes. ANPP, aboveground net primary production; BNPP, belowground net primary production, T, temperature.

Vulnerability of the Shortgrass Steppe Most of the simulated effects exhibited by the different scenarios can be explained in the context of how the model simulates nutrient and water limitation. Treatments that reduced nutrient stress (N) or water stress (CO2, +H2O) increased NPP. Increased temperature had little effect on simulated NPP because higher water stress is compensated for by lower nutrient stress resulting from enhanced N mineralization. Nitrogen addition decreased simulated belowground allocation as a fraction of total NPP because the model assumes that as N and water become more available, belowground allocation will decrease. However, the scenarios that decreased water stress (CO2, +H2O) did not lead to a decrease in belowground allocation (Fig. 15.3A). This occurred because the model assumes that roots must forage to supply additional N needed to support increased aboveground biomass that results from decreased water stress. These results suggest that the shortgrass

384

Ecology of the Shortgrass Steppe

steppe ecosystem is influenced by both water and N limitations, which interact on various scales to determine total NPP and biomass allocation. Increased N deposition was the only scenario to increase simulated system C substantially, even though the CO2 and increased precipitation treatments increased BNPP much the same as the increased N treatment. Contrary to our expectations, some scenarios showed an inverse relationship between NPP and system C. For example, decreased precipitation, the only scenario with lower NPP, also resulted in increased system C, whereas the CO2 and plus-water treatments both lost C (Fig. 15.2). This trend is a function of generally dry conditions in the shortgrass steppe, and the model assumption that microbial activity and NPP have different sensitivities to soil water content. The results of the DAYCENT simulations suggest that although increased water availability from elevated CO2 will increase NPP significantly (≈25%), the effect on soil C levels will be minor because increased decomposition rates approximately balance higher C inputs. If N deposition increases, DAYCENT predicts that NPP will increase by approximately 35% and soil C levels will increase significantly. One factor not considered in these simulations is the effect of elevated CO2 on species composition. Results from open-top chamber experiments performed at the shortgrass steppe (Morgan et al., 2004a) suggest that elevated CO2 favors drought-sensitive species with lower forage quality. Another limitation of these simulation experiments is that we varied only the amount of precipitation, but not the frequency of precipitation events. Recent DAYCENT simulations performed by Burke et al. (2002) show that decreasing the number of rainfall events while increasing the size of the events leads to substantially lower NPP, lower aboveground to belowground biomass and production ratios, and lower mineralization rates, but higher N gas emissions. Our DAYCENT simulations provide an excellent example of using a model to guide future experimental research. Our results indicate that key ecosystem processes in the shortgrass steppe are sensitive to climate change and elevated greenhouse gas levels. Future observations and experiments will be required to assess reliably how the shortgrass steppe will respond to changes in climate, CO2 levels, and N deposition.

This research was supported with funds from the NSF (LTER BSR9011 659, DEB 9632852), National Institute of Child Health and Human Development (#1 R01 HD33554), National Aeronautics and Space Administration (EOS NAGW2662), and Environmental Protection Agency (EPA Regional Assessment R824a39–01–0).

Acknowledgments

References Alward, R. D., J. K. Detling, and D. G. Milchunas.1999. Grassland vegetation changes and nocturnal global warming. Science 283:229–231.

The Shortgrass Steppe and Ecosystem Modeling 385 Antle, J. M., S. M. Capalbo, K. Paustian, and M. K. Ali. 2007. Estimating the economic potential for agricultural soil carbon sequestration in the central United States using an aggregate econometric–process simulation model. Climatic Change 80:145–171. Bledsoe, J. L., R. C. Francis, G. L. Swartzman, and J. D. Gustafson.1971. PWNEE: A grassland ecosystem model, pp. 179. In: U.S. IBP Grassland Biome technical report no. 64.Colorado State University, Fort Collins. Burke, I. C., E. C. Adair, R. L. McCulley, P. Lowe, S. Del Grosso, and W. K. Lauenroth. 2002. The importance of pulse dynamics in nutrient availability and ecosystem functioning: Evidence and questions. Presented at the Responses to Resource Pulses in Arid and Semi-arid Ecosystems Symposium, Tucson, Ariz., August 2002. Burke, I. C., T. G. F. Kittel, W. K. Lauenroth, P. Snook, C. M. Yonker, and W. J. Parton. 1991. Regional analysis of the central Great Plains: Sensitivity to climate variability. BioScience 41:685–692. Burke, I. C., W. K. Lauenroth, and W. J. Parton. 1997. Regional and temporal variation in aboveground net primary productivity and net N mineralization in grasslands. Ecology 78(5):1330–1340. Burke, I. C., D. S. Schimel, C. M. Yonker, W. J. Parton, L. A. Joyce, and W. K. Lauenroth. 1990. Regional modeling of grassland biogeochemistry using GIS. Landscape Ecology 4:45–54. Campbell, C. A., H. H. Janzen, K. Paustian, E. G. Gregorich, L. Sherrod, B. C. Liang, and R. P. Zentner. 2005. Carbon storage in soils of the North American Great Plains: Effect of cropping frequency. Agronomy Journal 97:349–363. Capinera, J. L., J. K. Detling, and W. J. Parton. 1983a. Assessment of range caterpillar (Lepidoptera: Saturniida) effects with a grassland simulation model. Journal of Economic Entomology 76:1088–1094. Capinera, J. L., W. J. Parton, and J. K. Detling. 1983b. Application of a grassland simulation model to grasshopper pest management on the North American shortgrass prairie, pp. 335–344. In: W. K. Lauenroth, G. V. Skogerboe, and M. Flug (eds.), Analysis of ecological systems: State-of-the-art in ecological systems. Elsevier, New York. Coffin, D. P., and W. K. Lauenroth. 1990. A gap dynamics simulation model of succession in a semiarid grassland. Ecological Modeling 49:229–266. Del Grosso, S. J., W. J. Parton, A. R. Mosier, M. D. Hartman, J. Brenner, D. S. Ojima, and D. S. Schimel. 2001. Simulated interaction of carbon dynamics and nitrogen trace gas fluxes using the DAYCENT model, pp. 303–332. In: M. Schaffer, L. Ma, and S. Hansen (eds.), Modeling carbon and nitrogen dynamics for soil management. CRC Press, Boca Raton, Fla. Del Grosso, S. J., W. J. Parton, A. R. Mosier, D. S. Ojima, A. E. Kulmala, and S. Phongpan. 2000. General model for N2O and N2 gas emissions from soils due to denitrification. Global Biogeochemical Cycles 14:1045–1060. Del Grosso, S. J., W. J. Parton, A. R. Mosier, M. K. Walsh, D. S. Ojima, and P. E. Thornton. 2006. DAYCENT national-scale simulations of nitrous oxide emissions from cropped soils in the United States. Journal of Environmental Quality 35:1451–1460. Eastman, J. L., R. A. Pielke, and D. J. McDonald. 1998. Calibration of soil moisture for large eddy simulations over the FIFE area. Journal of Atmospheric Science 55:1131–1140. Gifford, R. M., D. J. Barrett, and J. L. Lutze. 2000. The effects of elevated [CO2] on the C:N and C:P mass ratios of plant tissues. Plant and Soil 224:1–14. Hanson, J. D., J. W. Skiles, and W.J. Parton. 1988. A multi-species model of rangeland plant communities. Ecological Modeling 44:89–123.

386

Ecology of the Shortgrass Steppe

Holland, E. A., W. J. Parton, J. K. Detling, and D. L. Coppock. 1992. Physiological responses of plant populations to herbivory and their consequences for ecosystem nutrient flow. American Naturalist 140:685–706. Hu, S., F. S. Chapin, III, M. K. Firestone, C. B. Field, and N. R. Chiariello. 2001. Nitrogen limitation of microbial decomposition in a grassland under elevated CO2. Nature 409:188–191. Hunt, H. W., D. C. Coleman, C. V. Cole, R. E. Ingham, E. T. Elliott, and L. E. Woods. 1984. Simulation model of a food web with bacteria, amoebae, and nematodes in soil, pp. 346–352. In: M. J. Klug and C. A. Reddy (eds.), Current perspectives in microbial ecology. American Society for Microbiology, Washington, D.C. Hunt, H. W., M. J. Trlica, E. F. Redente, J. C. Moore, J. K. Detling, T. G. F. Kittel, D. E. Walter, M. C. Fowler, D. A. Klein, and E. T. Elliott. 1991. Simulation model for the effects of climate change on temperate grassland ecosystems. Ecological Modeling 53:205–246. Innis, G. S. 1978. Grassland simulation model: Ecological studies, 26. Springer-Verlag, New York. IPCC. 2001. Climate change 2001: The scientific basis–Contribution of Working Group I to the IPCC Third Assessment Report. Cambridge University Press, Cambridge, UK. IPCC. 2007. Intergovernmental panel on climate change. Climate change: The physical basis. Summary for policy makers. Contribution of Working Group I to the fourth assessment report of the IPCC. Online. Available at www.ipcc.ch. Kelly, R. H., W. J. Parton, M. D. Hartman, L. K. Stretch, D. S. Ojima, and D. S. Schimel. 2000. Intra- and interannual variability of ecosystem processes in shortgrass steppe. Journal of Geophysical Research: Atmospheres 10:20093–20100. Körner, C. 2006. Plant CO2 responses: An issue of definition, time and resource supply. New Phytologist 172:393–411. Lu, L., R. Pielke, G. Liston, W. J. Parton, D. Ojima, and M. Hartman. 2001. Implementation of a two-way interactive atmospheric and ecological model and its application to the central United States. Journal of Climate 14:900–919. McGill, W. B., H. W. Hunt, R. G. Woodmansee, J. O. Reuss, and K. H. Paustian. 1981. Formulation, process controls, parameters and performance of PHOENIX: A model of carbon and nitrogen dynamics in grassland soils, pp. 171–191. In: M. Frissel and J. Van Heem (eds.), Simulation of nitrogen behavior of soil–plant systems. Pudoc, Centre for Agricultural Publishing and Documentation, Wageningen, the Netherlands. Morgan, J. A., D. R. LeCain, A. R. Mosier, and D. G. Milchunas. 2001. Elevated CO2 enhances water relations and productivity and affects gas exchange in C3 and C4 grasses of the Colorado shortgrass steppe. Global Change Biology 7:451–466. Morgan, J. A., A. R. Mosier, D. G. Milchunas, D. R. LeCain, J. A. Nelson, and W. J. Parton. 2004a. CO2 enhances productivity, alters species composition, and reduces digestibility of shortgrass steppe vegetation. Ecological Applications 14:208–219. Morgan, J. A., D. E. Pataki, C. Körner, H. Clark, S. J. Del Grosso, J. M. Grünzweig, A. K. Knapp, A. R. Mosier, P. C. D. Newton, P. A. Niklaus, J. B. Nippert, R. S. Nowak, W. J. Parton, H. W. Polley, and M. R. Shaw. 2004b. Water relations in grassland and desert ecosystems exposed to elevated atmospheric CO2. Oecologia 140:11–25. Mosier, A. R., W. J. Parton, D. W. Valentine, D. S. Ojima, D. S. Schimel, and J. A. Delgado. 1996. CH4 and N2O fluxes in the Colorado shortgrass steppe, 1. Impact of landscape and nitrogen addition. Global Biogeochemical Cycles 10:387–399. Ojima, D. S., T. G. F. Kittel, T. Rosswall, and B. H. Walker. 1991. Critical issues for understanding global change effects on terrestrial ecosystems. Ecological Applications 1(3):316–325.

The Shortgrass Steppe and Ecosystem Modeling 387 Ojima, D. S., W. J. Parton, D. S. Schimel, and C. E. Owensby. 1990. Simulated impacts of annual burning on prairie ecosystems, pp. 118–132. In: S.L. Collins and L. L. Wallace (eds.), Fire in North American tallgrass prairies. University of Oklahoma Press, Norman, Okla. Parton, W. J., J. E. Ellis, and D. M. Swift. 1979. The impacts of strip mine reclamation practices: A simulation study, pp. 584–591. In: M. K. Wali (ed.), Ecology and coal resource development, vol. 12. Pergamon Press, New York. Parton, W. J., M. P. Gutmann, S. A. Williams, M. Easter, and D. Ojima. 2005. Ecological impact of historical land-use patterns in the great plains: A methodological assessment. Ecological Applications 15:1915–1928. Parton, W. J., M. D. Hartman, D. S. Ojima, and D. S. Schimel. 1998. DAYCENT: Its land surface submodel: Description and testing. Global Planetary Change 19:35–48. Parton, W. J., J. A. Morgan, G. Wang, and S. J. Del Grosso. 2007. Projected ecosystem impact of the prairie heating and CO2 enrichment experiment. New Phytologist.174:823–834. Parton, W. J., A. R. Mosier, and D. S. Schimel. 1988. Rates and pathways of nitrous oxide production in a shortgrass steppe. Biogeochemistry 6:45–58. Parton, W. J., and P. G. Risser. 1980. Impact of management practices on the tallgrass prairie. Oecologia 46(2):223–234. Parton, W. J., P. G. Risser, and R. G. Wright. 1980. Grazing responses in a prairie National Park. Environmental Management 4(2):165–170. Parton, W. J., D. S. Schimel, C. V. Cole, and D. S. Ojima. 1987. Analysis of factors controlling soil organic matter levels in Great Plains grasslands. Soil Science Society of America Journal 51:1173–1179. Parton, W. J., D. S. Schimel, and D. S. Ojima. 1994. Environmental change in grasslands: Assessment using models. Climatic Change 28:111–141. Parton, W. J., and R. C. Schnell. 1975. Use of an ecosystem model to study potential interactions between the atmosphere and the biosphere, pp. 721–726. In: Proceedings of the 1975 Summer Computer Simulation Conference, vol. I. Simulation Councils, La Jolla, Calif. Parton, W. J., J. M. O. Scurlock, D. S. Ojima, D. S. Schimel, D. O. Hall, M. B. Coughenour, E. Garcia Moya, T. G. Gilmanov, A. Kamnalrut, J. I. Kinyamario, T. Kirchner, T. G. F. Kittel, J. C. Menaut, O. E. Sala, R. J. Scholes, and J. A. van Veen. 1995. Impact of climate change on grassland production and soil carbon worldwide. Global Change Biology 1:13–22. Parton, W. J., J. S. Singh, and D. C. Coleman. 1978. A model of production and turnover of roots in shortgrass prairie. Journal of Applied Ecology 47:515–542. Parton, W. J., and F. M. Smith. 1974a. Exploring some possible effects of potential SSTinduced weather modification in a shortgrass prairie ecosystem, pp. 255–258. In: 6th Conference on Aerospace and Aeronautical Meteorology. American Meteorological Society, Boston, Mass. Parton, W. J., and F. M. Smith. 1974b. Simulating the effects of growing season rainfall enhancement and hail suppression on the production, consumption, and decomposition functions of a native shortgrass prairie ecosystem, pp. 523–528. In: 4th Conference on Weather Modification. American Meteorological Society, Boston, Mass. Pepper, D. A., S. J. Del Grosso, R. E. McMurtrie, and W. J. Parton. 2005. Simulated carbon sink response of shortgrass steppe, tallgrass prairie and forest ecosystems to rising [CO2], temperature and nitrogen input. Global Biogeochemistry Cycles 19: GB1004, doi:10.1029/2004GB002226. .

388

Ecology of the Shortgrass Steppe

Pielke, R. A, Sr., G. E. Liston, L. Lu, P. L. Vidale, R. L. Walko, T. G. F. Kittel, W. J. Parton, and C. B. Field. 1997. Coupling of land and atmospheric models over the GCIP area: CENTURY, RAMS, and SiB2C. Presented at the 13th annual Conference on Hydrology. 77th AMS annual meeting, Long Beach, Calif., February 2–7. Pielke, R. A., T. Stohlgren, W. Parton, J. Moeny, N. Doesken, L. Schell, and K. Redmond. 2000. Spatial representativeness of temperature measurements from a single site. American Meteorological Society 81(4):826–830. Schimel, D. S., B. H. Braswell, E. A. Holland, R. McKeown, D. S. Ojima, T. H. Painter, W. J. Parton, and A. R. Townsend. 1994. Climatic, edaphic, and biotic controls over storage and turnover of carbon in soils. Global Biogeochemical Cycles 8:279–293. Schimel, D. S., T. G. F. Kittel, and W. J. Parton. 1991. Terrestrial biogeochemical cycles: Global interactions with the atmosphere and hydrology. Tellus 43AB:188–203. Schimel, D. S., W. J. Parton, T. G. F. Kittel, D. S. Ojima, and C. V. Cole. 1990. Grassland biogeochemistry: Links to atmospheric processes. Climatic Change 17:13–25. Vidale, P. L., R. A. Pielke, A. Barr, and L. T. Steyaert. 1997. Case study modeling of turbulent and mesoscale fluxes over the BOREAS region. Journal of Geophysical Research 102:29167–29188. Walko, R. L., L. E. Band, J. Baron, T. G. F. Kittel, R. Lammers, T. J. Lee, D. S. Ojima, R. A. Pielke, C. Taylor, C. Tague, C. J. Tremback, and P. L. Vidale. 2000. Coupled atmosphere–biophysics–hydrology models for environmental modeling. Journal of Applied Meteorology 39:931–944. Woodward, F. I. 2002. Potential impacts of global elevated CO2 concentrations on plants. Current Opinion Plant Biology 5:207–211.

16 Effects of Grazing on Vegetation Daniel G. Milchunas William K. Lauenroth Ingrid C. Burke James K. Detling

Evolutionary History of Grazing and Semiaridity Grazing by large native ungulates and semiaridity are the two main forces that have had a large influence in shaping the current-day structure of the shortgrass steppe ecosystem (Milchunas et al., 1988). With the uplift of the Rocky Mountain chain during the Miocene (approximately one million years ago), forests of the Great Plains were gradually replaced by grasslands (Axelrod, 1985). Large grazing and browsing animals inhabited the Great Plains during the middle to late Pleistocene, as did grasses of the genera Stipa, Agropyron, Oryzopsis, and Elymus (Axelrod, 1985; Stebbins, 1981). Bison occurred both east and west of the Rockies during the Wisconsin glacial period in the latter part of the Pleistocene (Wilson, 1978). During the early Holocene, approximately 10,000 years ago, bison and grasses of the genera Bouteloua, Buchloë, Andropogon or Schizachyrium, and Sorghastrum concomitantly increased throughout the Great Plains (Stebbins, 1981), but bison did not proliferate west of the continental divide (Mack and Thompson, 1982; Van Vuren, 1987). The natural shift in fauna from horses, pronghorn, and camels to bison and wild sheep from Eurasia is thought to have favored the spread of shortgrasses such as Bouteloua and Buchloë (Stebbins, 1981). Furthermore, grassland flora east and west of the Rocky Mountains probably had separate origins (Leopold and Denton, 1987). The shortgrass steppe is unique from other North American semiarid ecosystems in having bison play an important role. Bison did not proliferate west of the Rocky Mountains as they did on the Great Plains to the east. This is due in part to a lack of coincidence in timing of bison lactation and the phenological development of C3 grasses in the more Mediterranean–like climate west of the Rockies, in contrast to 389

390

Ecology of the Shortgrass Steppe

the mix of C3 and C4 grasses and pattern of spring–summer precipitation on the Great Plains (Mack and Thompson, 1982). Other explanations for the low numbers of bison west of the Rocky Mountains include physiographic barriers restricting immigration (Kingston, 1932), low protein content of forage (Daubenmire, 1985; Johnson, 1951), heavy snowfall as a cause of mortality (Daubenmire, 1985), and low aboveground primary production coupled with disjunct suitable habitat (Van Vuren, 1987). Bison also did not prosper in the southwestern United States, nor did a large herbivore fauna develop in South America (Stebbins, 1981). The evolution of plains grasses was rapid in North America compared with South America (Stebbins, 1981). Evolutionary lines of bison in North America included six species, three of which were endemic, and the extant Bison bison subsp. bison of the plains that developed about 4000 to 5000 years ago (McDonald, 1981). The shortgrass steppe occupies the driest portion of the Great Plains, because of its location in the rain shadow created by the Rocky Mountains (Pielke and Doeskin, chapter 2, this volume). Precipitation ranges from 300 mm in the west to 550 mm in the east (Lauenroth and Milchunas, 1991). This trend of increasing precipitation continues from west to east and from shortgrass steppe to mixedgrass prairie to tallgrass prairie to eastern deciduous forest. Across this gradient, there is increased allocation to aboveground plant production, and a shift in the relative importance of belowground competition for soil water on the dry end to more intense competition for nitrogen and light in a dense canopy on the wet end. Although approximately 90% of plant biomass in the shortgrass steppe is belowground, BNPP is only 67% of total ANPP (Milchunas and Lauenroth, 1992, 2001). Grazed or not, the plant canopy in this dry environment is short and sparse relative to many other grasslands (Milchunas et al., 1988). Plants of the shortgrass steppe are, therefore, unique in the western hemisphere in terms of having evolved both with a long evolutionary history of grazing by large herbivores and with semiaridity. Selection for individuals in a plant population that have characteristics providing a greater tolerance or avoidance of grazing can be rapid. Various studies have indicated phenotypic changes within 4 months (Brougham and Harris, 1967) and apparent genotypic changes in population structure in 10 to 15 years (Detling and Painter, 1983; Jaramillo and Detling, 1988), 13 years (Peterson, 1962), 25 years (Briske and Anderson, 1992), and 35 years (Kemp, 1937). Thus, although the Great Plains flora is considered young and there are few endemic plants, insects, and birds (Mengel, 1970; Ross, 1970; Stebbins, 1981), the 9000- to 10,000-year association between grasses and herbivores has been sufficient to separate it distinctly from other ecosystems in the western hemisphere, particularly in terms of its capacity to withstand herbivory (Larson, 1940; Mack and Thompson, 1982; Milchunas et al., 1988). Although fire has been an important force maintaining and shaping the structure of the more productive eastern grasslands of the Great Plains, it has not been as important in the shortgrass steppe with its low fuel loads. Weaver and Clements (1938) considered the shortgrass steppe to be a disclimax community, caused by the disturbance of grazing by domestic animals. Although grazing can shift mixed-grass prairie to resemble shortgrass steppe more closely (Lauenroth et al., 1994), Larson (1940) argued that bison maintained the shortgrass steppe, and that it should, therefore,

Effects of Grazing on Vegetation 391

not be considered a disturbance community. However, early classifications considering all of the Great Plains as mixed-grass prairie (Weaver and Albertson, 1956) may have led to confusion; responses to grazing differ across shortgrass, mixed-grass, and tallgrass types (Lauenroth et al., 1994). Although all grasslands of the Great Plains coevolved with Bison, we now know that the shortgrass steppe is climatically maintained (Lauenroth, chapter 5, this volume; Lauenroth and Milchunas, 1991). Grazing and semiaridity can be convergent, complementary selection forces (Milchunas et al., 1988), contributing to the unique responses to grazing that we will examine in this chapter.

Recent History of Land Use in the Shortgrass Steppe Estimates of the numbers of bison that occupied the plains prior to their extinction as a commercially, socially, and ecologically important species in the mid 1870s range from 5.5 to 60 million, based upon extrapolation of observations by early explorers and hide market numbers or various combinations of potential carrying capacity and early observations (Hart, chapter 4, this volume; Shaw, 1995). Some single herds were estimated to have totaled four million (Hornaday, 1889), and others covered 3500 sq. km. (Farnham, 1839 [as cited in McHugh, 1972]). The demise of the bison coincided with a period (1870–1890) of large increases in the numbers (tripling) of cattle in the United States (U.S. Department of Commerce, Bureau of Census, 1935–1982 [as cited in Joyce, 1989]). Up to about 1900, livestock grazed freely over the shortgrass steppe and received little or no supplementation from crop or hay meadows (Cook and Redente, 1993). Most ownership boundaries were established by fencing by 1900, and permits for grazing of public lands were being issued in the early 1900s. By 1930, cropland had become an important component of the landscape, and supplementation increased to about 20% of feed supply. Grazing of public lands was further regulated with the passing of the Taylor Grazing Act in 1934 (Bement, 1993). Cultivated cropland reached a peak prior to the Dust Bowl drought of 1934 to 1937, after which much of the abandoned land reverted back to public ownership. Current land use in the shortgrass steppe region is approximately 70% rangeland and 30% cropland (Lauenroth et al., 1994; Parton et al., 2003, Hart, chapter 4, this volume). All public and private grassland is grazed by domestic livestock, except for small areas set aside as experimental or “nature” areas. Some grassland is former cropland that was abandoned in the 1930s; approximately 20% to 30% of the Pawnee National Grassland (PNG) and the Central Plains Experimental Range (CPER) is in old-field succession following cessation of cultivation, and is still ecologically (and visually) discernible from native grassland (Peters et al., chapter 6, this volume). Fluctuations in land use have been driven by economics, government programs, and weather cycles (Joyce, 1989). However, basic environmental constraints have not allowed cropland area expansion since the 1950s (Parton et al., 2003). Much of the shortgrass steppe is marginal for sustaining crop agriculture, is highly susceptible to erosion when the native perennial grass cover is destroyed, and recovers very slowly from disturbance.

392

Ecology of the Shortgrass Steppe

Herbivory and Grazing in the Shortgrass Steppe The Herbivore Pathway of Energy Flow and Grazing Intensities For moderately stocked shortgrass steppe, aboveground consumption by all trophic levels is dominated by cattle (Table 16.1 [Lauenroth and Milchunas, 1991]). Although cattle are the most obvious herbivore on the shortgrass steppe, their contribution to energy flow accounts for only 13% of total nonsaprophytic consumption. Consumption in terms of herbivory is approximately equally divided among cattle (aboveground), arthropods (both above- and belowground), and nematodes (belowground), with lagomorphs, rodents, and birds each totaling less than 1%. Although cattle do not dominate total herbivory (because of the large amount of belowground herbivory by arthropods and nematodes), cattle are the dominant aboveground herbivore. Herbivory by cattle is a dominant force on plant communities from aboveground, but their role as a dominant herbivore diminishes when placed in the context of the total plant. Herbivory in grasslands and shrublands is large in comparison with most other systems. In forests, folivory is typically 10% or less of annual foliage standing crop of trees, except during periods of insect outbreaks (Detling, 1989; Schowalter et al., 1986). Herbivory by native ungulates on Serengeti grasslands can reach 92% removal (McNaughton, 1985), which is greater than a heavily grazed shortgrass steppe. Light, moderate, and heavy grazing of a shortgrass steppe is generally considered to be 20%, 40%, and 60% removal of ANPP, respectively (Klipple and Costello, 1960). What is considered heavy grazing in some grasslands of the Great Plains may be considered moderate in the shortgrass steppe. For example, Heitschmidt et al. (1985) considered heavy grazing of mixed-grass prairie in Texas to average approximately 40% to 50% consumption of ANPP. Table 16.1 The Importance of Different Groups of Consumers in Terms of Herbivorous Processing of Aboveground, Belowground, and Total Plant Energy in the Northern Shortgrass Steppe Herbivory

Consumer Group Ruminants Lagomorphs Rodents Birds Arthropods Macro Micro Nematodes Total

Total NPP, 4.77 0.06 0.04 0.01 5.05 4.20 0.85 5.11 15.04

Aboveground NPP, % 29.90 0.39 0.22 0.06 5.11 5.11 — — 35.68

Belowground NPP, % — — — — 5.04 4.03 1.01 6.08 11.12

Values show the proportion of total (NPP), aboveground (ANPP), and belowground (BNPP) net primary production that is consumed by the group of herbivores indicated. (Adapted from Lauenroth and Milchunas [1991].)

Effects of Grazing on Vegetation 393

Stocking rates in different systems are generally adjusted to reflect the capacity of the vegetation to withstand the grazing pressure, and there are large differences across the world between the qualitative descriptors of grazing intensity and the quantitative estimates of consumption (Milchunas and Lauenroth, 1993). Average forage production for the shortgrass steppe is approximately 70 g · m–2 · y–1 in moderately grazed level uplands (Fig. 16.1; see Lauenroth et al., chapter 12, this volume) compared with an average of 285 g · –2 · y–1 in the moderately grazed

26

Light grazing 20 Mean 14 8 2

# of occurrences

26

Mean

Moderate grazing

20 14 8 2 26

Heavy grazing 20 14

Mean

8 2 25 30 35 40 50 60 70 80 90 100 110 120 130 140 >140

Forage production (g m-2 yr-1) Figure 16.1 Distributions (number of annual occurrences) of forage production (measured in grams per square meter per year) in long-term lightly, moderately, and heavily grazed treatments in the northern shortgrass steppe. Data are for 89, 97, and 88 site samplings for the aforementioned respective treatments, and span the years 1940 to 1990. (From Milchunas et al. [1994].)

394

Ecology of the Shortgrass Steppe

sites of the Texas mixed-grass prairie (Heitschmidt et al. 1985). Across grasslands of the Great Plains, there is an inverse relationship between grazing intensities and ANPP within a particular qualitative description of moderate or heavy grazing (Fig. 16.2). Impacts of grazing on plant species composition and ANPP generally increase with increasing ANPP of a vegetation type (Milchunas and Lauenroth, 1993). Most of the studies we report on here were conducted at the CPER and the PNG. Grazing at the CPER is primarily on a summer–winter pasture basis.

80

Qualitative Description of Grazing: Moderate (r2 = 0.46)

60

Quantitative Grazing Intensity (% ANPP consumed)

M

M

40

M MMM

MM M M

M

M M M MM

M M M M M M M M M

M

20

M

0

80

H H

60

H H

H HH H

H

HH

H

HH

H H

H

H

40

HH H

Qualitative Description of Grazing: Heavy (r2 = 0.51)

20

0 0

100

200

300

H

H

400

500

Aboveground Net Primary Production (g m-2 yr-1) Figure 16.2 Quantitative grazing intensity (measured percentage of aboveground net primary production [ANPP] consumed) in relation to ANPP (measured in grams per square meter per year) for various authors’ qualitative description of moderate and heavy grazing intensities. Studies are from grasslands of the Great Plains. (Compiled from data found in Appendix I of Milchunas and Lauenroth [1993].)

Effects of Grazing on Vegetation 395

The summer grazing season begins in late May, approximately 1 month after the start of the cool-season growth period, and extends for a maximum of 184 days (late November). The length of the grazing period depends upon the time necessary to achieve the intensity criteria, and has been as short as 48 days under the heavy grazing regime (Ashby et al., 1993; Hart and Ashby, 1998). The frost-free dates are from May 15 through September 15, but vegetation is often brown by August and short green-ups and periods of growth depend upon fall precipitation. Since 1939, numbers of animals on the half-section pastures have ranged from 6 to 22, 11 to 29, and 14 to 45 yearling heifers under the light, moderate, and heavy stocking regimes, respectively (Ashby et al., 1993). Ungrazed exclosures were established in 1939 (Fig. 16.3 a, b). Although an average removal of 20%, 40%, and 60% of ANPP for lightly, moderately, and heavily grazed conditions, respectively, has been maintained since 1939, the means of achieving these removal rates has changed. From 1939 until 1964, intensity criteria were based upon the percent of ANPP removed, but since 1965 the intensity criteria have been based upon leaving a certain amount of biomass. Residual biomass of 22, 34, and 50 g · m–2 has been the goal for lightly, moderately, and heavily grazed areas, respectively (Ashby et al., 1993; Hart and Ashby, 1998). Across years, the average percent removal of ANPP remains approximately the same under the two methods, but there are several important differences between them. From a management standpoint, it is easier for both researchers and ranchers to estimate residual biomass than to predict actual production to allow consumption of a certain percentage of that amount. From an ecological standpoint, the two methods are opposite in terms of conditions under which year-to-year grazing pressures on plant communities are most severe. In years of low productivity, removing a percentage of a small amount of growth may have deleterious effects; whereas during highly productive years, removing a percentage still leaves a large residual. The opposite occurs when managing based on residual. Defoliation pressure will be lessened under conditions of drought stress. Under extreme drought, for example, the required amount of residual to be left may never be produced, so animals would not even be allowed to graze. During years of high plant production, grazing pressures are higher when managing based on residuals. Thus, leaving a certain amount compared with consuming a certain amount lessens the grazing pressure in years of low productivity, but increases the grazing pressure in years of high productivity. This tends to lower year-to-year variability in the impact of unfavorable growing conditions on the system, but it also means greater year-to-year fluctuation in potential numbers of animals allowed on the range. Plant Communities, Grazing Behavior, Diet Selection, and Forage Quality In this section we briefly describe the shortgrass steppe landscape and the manner in which domestic livestock forage at various scales. The interactions between wildlife, other consumers, and domestic livestock are addressed separately by Milchunas and Lauenroth in chapter 18, and the livestock responses to the grazing

Figure 16.3 The shortgrass steppe is dominated by prostrate C4 grasses across the gently rolling topography. (A) Loamy uplands ridgetop site at the CPER that has been heavily grazed since 1939 (left of fence) or ungrazed since 1939 (right of fence). (Photo by

Effects of Grazing on Vegetation 397

treatments and effects of grazing management systems are discussed by Hart and Derner in chapter 17. As described by Lauenroth in chapter 5, most plant communities in the shortgrass steppe are dominated by B. gracilis, but there are exceptions. The dominant species and other warm-season (C4) grasses comprise a large proportion of the basal cover and of the ANPP in upland communities. Within the upland communities, Senft et al. (1985) identified six community types based upon the relative proportions of the important grass, half-shrub, and cactus species, as well as soil texture and topographic (catenary) position. In lowland areas of the CPER, Atriplex canescens shrublands occupy bands of sandy soils (overflow sites) along larger ephemeral stream channels (see Fig. 18.1B, Milchunas and Lauenroth, chapter 18, this volume), but understory species are similar to those of level upland habitat (Liang et al., 1989). Because low abundance of cool-season (C3) species is a limiting factor in animal production in the shortgrass steppe region (Hart, chapter 4, this volume), the C4 shrub A. canescens habitat is often used as winter pasture. Herbivore foraging patterns are based upon decisions that the animal must make at several levels of spatial resolution, and these can vary through time (Senft et al., 1987). Landscape-scale foraging patterns in the shortgrass steppe can be categorized according to topographic or soil characteristics, which are associated with differences in plant communities and primary productivity (Lauenroth, chapter 5, this volume). Diet selection is a function of the chemical and physical characteristics of plant species and plant parts (palatability), and selection of feeding areas is often related to diet preference (Senft, 1989). Diet selection by cattle can directly affect plant community composition resulting from differences in species’ abilities to tolerate or avoid herbivory, or indirectly through altering competitive relationships between neighboring plants. Differential habitat use can affect the distributions of communities as well as the degree of compositional change that may occur in both primary producer and consumer populations. Although considered to be large generalist herbivores, cattle display a reasonable degree of diet selectivity (Table 16.2). In some studies, the dominant grass (B. gracilis) appears to be consumed in proportion to its cover (Senft et al., 1984a), but other cattle diet studies have found its consumption to be smaller than its proportion of cover (Hanson and Gold, 1977; Shoop et al., 1985; Vavra et al., 1977). Agropyron smithii, a major component of mixed-grass prairie, has been found to be highly preferred (Lauenroth and Milchunas, 1991); it is also an important component of small-mammal diets. The dominant upland subshrub (Artemisia frigida) and forb (Sphaeralcea coccinea) are usually consumed in lesser amounts

Daniel G. Milchunas.) (B) Loamy uplands swale site at the CPER that has been heavily grazed (left of fence) or ungrazed since 1939 (right of fence). (Photo by Daniel G. Milchunas.) (C) Grazing lawn-type physiognomy in heavily grazed swale with taller vegetation inside cactus clumps where cattle do not graze. (Photo by Mark Lindquist.) Note that in (A) and (B), long-term ungrazed vegetation is not dramatically taller than long-term heavily grazed vegetation, but differences inside and outside the exclosure are greater in the more productive swales than in the ridgetops.

398

Ecology of the Shortgrass Steppe

Table 16.2 Percentages of Major Plant Species Available to and Consumed by Cattle, and Their Preference Rankings in Long-Term Moderately Stocked Pastures in the Northern Shortgrass Steppe Species Psoralea tenuiflora Stipa comata Agropyron smithii Chenopodium album Eriogonum effusum Gutierrezia sarothrae Sporobolus cryptandrus Astragalus gracilis Leucocrinum montanum Carex eleocharis Gaura coccinea Lepidium densiflorum Sitanion hystrix Bouteloua gracilis Aristida longiseta Chrysothamnus nauseosus Astragalus spatulatus Sphaeralcea coccinea Buchloë dactyloides Plantago patagonica Artemisia frigida Lappula redowskii Vulpia octoflora Chenopodium leptophyllum Chrysopsis villosa Euphorbia glyptosperma Mirabilis linearis Picradeniopsis oppositifolia

Bites, %

Cover, %

Preference, %a

0.19 3.13 4.96 1.05 0.33 0.72 3.96 0.08 0.08 9.95 0.06 2.36 4.55 59.27 1.16 0.25 0.33 1.91 4.77 0.14 0.22 0.03 0.03 0.03 0.00 0.00 0.00 0.00

0.03 0.51 0.86 0.23 0.09 0.21 1.32 0.03 0.03 5.24 0.03 1.35 3.96 59.34 1.78 0.49 0.67 3.88 10.67 0.61 1.10 0.20 0.41 0.50 0.25 0.20 0.12 0.29

6.63 6.08 5.77 4.50 3.79 3.42 3.01 2.84 2.84 1.90 1.90 1.75 1.15 1.00 0.65 0.51 0.49 0.49 0.45 0.23 0.20 0.14 0.07 0.06 0.00 0.00 0.00 0.00

a Percent bites/percent cover. Values based upon 3603 bites from belt transects in six sites over 3 years (Milchunas, unpublished data).

than their abundance. Seven grass or Carex species comprised the greatest proportion of cattle diets in moderately stocked grassland. However, other cattle diet studies at the CPER site found much less consumption of B. gracilis (Hanson and Gold, 1977; Shoop et al., 1985; Vavra et al., 1977). Not unexpectedly, dietary preferences of cattle change with time of year. Schwartz and Ellis (1981) observed that C4 warm-season grasses increased and C3 cool-season grasses decreased in importance in cattle diets from spring through late fall. Forbs were important items in the diet in summer whereas shrubs were selected primarily in spring. The major shrub Atriplex was an important food source in winter and early spring both in terms of quantity (up to 55% in March) and nutrient content (Cook et al., 1977; Shoop et al., 1985). An intensive study of grazing behavior in moderately grazed upland plains indicated that cattle may select foraging areas by plant community type or

Effects of Grazing on Vegetation 399

topographic zones, although these two classifications are not independent in many areas (Senft et al., 1985). Preference for plant communities and topographic zones changed seasonally. Lowland swales and draws were heavily utilized during the growing season, as were fence lines and areas adjacent to watering tanks (Table 16.3). In contrast, ridgetops and south-facing slopes were more heavily utilized during the dormant season. Senft et al. (1985) hypothesized that the seasonal shift was the result of movement to less preferred areas after growth ceased and forage became depleted in the preferred areas. Simulations have suggested that increasing stocking rates would increase use of less productive communities because of a more rapid depletion of preferred areas (Senft, 1989). Increasing stocking rates was predicted to have a much greater effect on species preferences in the diet than on preferences for plant communities/topographic zones (Fig. 16.4). Utilization of lowland swales under heavy stocking rates decreases plant species diversity that is normally associated with gradients in soil quality (Milchunas et al., 1989). Relative preferences for topographic zones/plant communities may also vary over long timescales of grazing pressure. Varnamkhasti et al. (1995) found greater quantities of forage removal from lowland swales in long-term lightly grazed pastures than in heavily grazed pastures, and could not attribute this pattern to differences in size of swales or distances to water in the two treatments. Forage quality may also interact with topographic location and grazing intensity. Nitrogen availability (standing nitrogen) was greater in the lowland swales of lightly grazed than of heavily grazed treatments (Milchunas et al., 1995),

Table 16.3 Grazing Preferences for Plant Communities and Topographic Zones for Two Seasons in Upland Plains Sites in the Northern Shortgrass Steppe Plant Community

Period Growing season (Apr–Oct) Dormant season (Nov–Mar)

Buda-Bogr

Buda-AgsmCarex

Agsm-Dist

Bogr-Oppo

Bogr-ErefOppo

Bogr-Eref

1.10

1.68

1.11

0.60

0.72

1.39

1.55

0.94

0.70

1.05

1.38

0.40

Topographic Zone

Period Growing season (Apr–Oct) Dormant season (Nov–Mar)

Ridgetops

South-Facing North-Facing Draws and Slopes Slopes Swales

Fence Lines

Watering Area

0.43

0.93

0.86

1.39

1.05

1.95

1.61

1.16

1.09

0.80

0.20

1.53

Values represent the amount of time spent grazing compared with the area available. (Adapted from Senft et al. [1985].). Agsm, Agropyron smithii; Bogr, Bouteloua gracilis; Buda, Buchloë dactyloides; Dist, Distichlis stricta; Oppo, Opuntia polyacantha; Eref, Eriogonum effusum.

(A) 100

Diet Composition (%)

80

60 40 20

0 0

1

0.5

2

1.5

2.5

Relative Stocking Rate Agropyron smithii Carex spp. Sphaeralccea coccinea

B. gracilis & B. dactyloides Other grasses Forbs and shrubs

Foraging time (%)

(B) 100

Communities

80

C B

60

F A

40

E

20

D

0 0

0.5

1

1.5

2

2.5

Relative stocking rate Figure 16.4 (A–B) Simulations of the effect of stocking rate on grazing behavior of cattle at plant community and plant species levels of selection in the shortgrass steppe. Relative stocking rate effects on species composition of diets (percent) (A), percent of time spent foraging in different plant communities (B), relative preferences for species or plant groups.

400

Relative preferences

(C) 10

A. smithii Carex spp. S. coccinea gracilis & {B. B. dactyloides Other grasses Forbs & Shrubs

1

0.1 0

1

0.5

2.5

2

1.5

Relative stocking rate

Relative preferences

(D) 10

Community A B C

D E F

1

0.1 0

0.5

1

1.5

2

2.5

Relative stocking rate Figure 16.4 (C–D) Simulations of the effect of stocking rate on grazing behavior of cattle at plant community and plant species levels of selection in the shortgrass steppe. Relative stocking rate effects on relative preferences for species (C), and relative preferences for six plant communities in the pasture (D). The plant communities are A, Buchloë dactyloides–Bouteloua gracilis; B, B. dactyloides–Pascopyrum smithii–Carex spp.; C, P. smithii–Distichlis spicata; D, B. gracilis–Opuntia polyacantha; E, B. gracilis–Eriogonum effusum–O. polyacantha; F, B. gracilis–E. effusum. (From Senft [1989].)

401

402

Ecology of the Shortgrass Steppe

suggesting that long-term preferential grazing of swales may have reduced time spent in these areas in the heavily grazed treatments relative to those in the lightly grazed treatments. Senft et al. (1985) found that standing nitrogen in preferred species (r = .75), biomass of preferred species (r = .71), standing nitrogen in live plants (r = .71), and standing live biomass (r = .69) were all correlated with community/topographic zone preferences, but that total aboveground biomass was only weakly correlated (r = .45) with preference. Nutrient density (quality × quantity) appears to be an important factor in plant community preference by cattle, particularly in the shortgrass steppe, where sparse canopy cover and prostrate growth forms mean that overall nutrient density is low and bite sizes are small. Similar to differences in plant communities, forage or diet quality display less seasonal variation in the shortgrass steppe than in many other systems. This may be due in part to the high-quality, dormant-season curing characteristics of the dominant species (B. gracilis) and to the low stem-to-leaf ratios of this very short-stature grass. These seasonal dynamics are evident even within Colorado shortgrass–mixed-grass types (Fig. 16.5). Tall-stature grasses generally display greater declines in nutritive value with advancing phenology (Fig. 16.6). On the other hand, although shrubs and forbs maintain greater protein, carotene, and phosphorus levels with advancing maturity (Cook et al., 1977), these are not large components of most shortgrass steppe plant communities. Crude protein and digestible dry matter of cattle diets vary little more with season of the year than between plant communities on the shortgrass steppe (Senft et al., 1984a). Both nitrogen and energy budgets of cattle can fall below maintenance requirements by late in the year when supplementation is not provided (Table 16.4) (Senft et al., 1984b). These nutritional factors have implications for cattle weight gains and

Digestible Protein in Cattle Diet (average %)

12

Shortgrass steppe Mixed grass prairie Bunchgrass

10 8 6 4 2

J

F

M

A M

J J Month

A

S

O

N D

Figure 16.5 Seasonal digestible protein (average percent) in cattle diets for the shortgrass steppe in southeastern Colorado, a mixed-grass type in sandy soils (Akron, Colorado), and bunchgrass range in northwestern Colorado. (Data from Cook et al. [1977].)

15

Vegetative Head Seed Shatter

Protein (%)

13 11 9 7 5 3 1

BOGR BUDA

AGSM KOCR

ANSC ANGE

Short grasses

Mid grasses

Tall grasses

Figure 16.6 Protein concentrations of short-, mid-, and tallgrass species at vegetative, head, seed, and shatter phenological stages of growth (generally spring through early winter periods). AGSM, Agropyron smithii; ANGE, Andropogon gerardi; ANSC, A. scoparius; BOGR, Bouteloua gracilis; BUDA, Buchloë dactyloides; KOCR, Koeleria cristata. These are important species in the shortgrass steppe, mixed-grass prairie, and tallgrass prairie. The upper dashed line is the mean for the vegetative phenological stage of shortgrasses and the lower dashed line is the mean for the shatter phenological stage for shortgrasses. (Data from Cook et al. [1977].)

Table 16.4 Seasonal Nitrogen and Metabolizable Energy Required and Eaten (Intake) for Cattle Grazing at Moderate Stocking Rates in Upland Plains Sites in the Northern Shortgrass Steppe Nitrogen, g · d–1 Month May June July August September October November January March a

ME, Mcal · d–1

Body Weight, kg

Requireda

Intake

Requireda

Intake

56 56 64 64 64 74 74 74 74

89 139 97 112 102 97 72 69 71

8.2 8.2 9.4 9.4 9.4 10.6 10.6 10.6 10.6

9.6 14.8 14.7 16.8 14.9 18 16.1 9.8 6.7

National Research Council (1976). Adapted from Senft et al. (1984b).

245 252 287 292 291 331 326 323 317

404

Ecology of the Shortgrass Steppe

management practices (Hart, chapter 4, this volume), as well as for the impacts of livestock grazing on wild herbivores (Milchunas and Lauenroth, chapter 18, this volume).

Effects of Grazing on Vegetation Plant Community Structure It is well known that different methods of sampling vegetation may bias results with respect to species or life-form abundance and composition (Milchunas, 2006). This is especially true for grazing studies in which animals are selectively removing plant parts through the season or year. Measures such as canopy cover or biomass of species, unless they are from temporarily caged plots, confound current-year removal by grazing with potential long-term effects on population dynamics (mortality, establishment) and plant growth capacity (potential productivity). Current-year removal always has an impact on community structure, which may or may not manifest in long-term effects if the removals are terminated. Other measures such as species density, frequency, or basal cover do not confound current-year removal with long-term population dynamics, but do have other drawbacks. It is important to keep in mind what the technique measures when interpreting studies that examine the effects of grazing on plant communities. Species, Life-Forms, and Functional Groups The effects of grazing on plant community structure have been studied at several northern and southern shortgrass steppe sites. The longest and most intensively studied controlled experiments are at the CPER in the northern shortgrass steppe. Based upon periodic sampling from 1940 through the early 1950s, researchers concluded that the shortgrass steppe shows little response to grazing (Klipple and Costello, 1960). Changes in species abundances resulting from fluctuations in weather were much greater than those resulting from grazing. This prompted the authors to comment that with heavy grazing “some changes in the vegetation, such as the sod-type growth of B. gracilis (blue grama) and the disappearance of highly palatable species, would go unnoticed if similar ranges subject to other grazing treatments were not nearby for comparison” (p. 78). The shift to prostrate growth forms of B. gracilis, and the sodlike appearance that Harper (1969) and McNaughton (1984) would later term grazing lawns (Fig. 16.3C), was considered an undesirable consequence of heavy grazing. Change in morphology greatly reduced available forage to the grazing animals, cattle weight gains were lower, and animal mortality increased significantly (Fig. 16.7). Impacts of heavy grazing on plant community structure were considerably lower than impacts on cattle, however. This has potentially important ramifications for the shortgrass region, where much of the land is owned by private ranchers, and economics play an important role in management decisions. Loss of profits resulting from

Cattle weight gain

6 4 2 0

1 through 3 8

(C)

6 4 2 0

1 through 5

6 through 10

Years of Grazing Treatments light

(B)

130 100 70 40 10

7 through 9

1 through 3 Plant species similarity

(g m-2)

8

(kg head-1 6 months-1)

(A)

(Whittaker Index Association)

10

(# period-1)

Cattle mortality

B. gracilis Production (1% cover) -1

Effects of Grazing on Vegetation 405

1.0

7 through 9 (D)

0.8 0.6 0.4 0.2 0

1 through 3

7 through 9

Years of Grazing Treatments

Grazing intensity moderate

heavy

Figure 16.7 Vegetation and cattle responses to grazing intensity treatments in the northern shortgrass steppe averaged for the first 3 years of treatment (1940–1942) and years 7 through 9 (1945–1949). (A) Bouteloua gracilis yield per unit area of ground cover. (B) Cattle weight gain. (C) Cattle mortality.(D) Plant community species similarity (ungrazed compared with light, moderate, and heavily grazed treatments calculated using the Whittaker [1952] index of community association with cover data). (Compiled from data in Klipple and Costello [1960].)

overgrazing are important compared with shifts in plant community composition, and an animal/profitability threshold may be reached before the system moves to an alternative state. In other words, the effects on the animals are greater than the effects on plant community species composition. Furthermore, the increased vegetative cover with heavy compared with light grazing has effects on plant community structure that are discussed later. Still, some changes in shortgrass steppe species abundances do occur with changes in grazing intensity, and our understanding of the regulating factors and specifics of population dynamics has progressed with additional studies and years of treatment. Results from early years did not show much difference between grazing treatments in cover of the dominant B. gracilis (Table 16.5). Cover of B. dactyloides, a stoloniferous shortgrass, increased with grazing, as did the annual Festuca octoflora. Midheight grasses such as S. comata and the palatable A. smithii decreased with grazing.

Table 16.5 Species Abundance or Composition of Long-Term Grazing Treatments in the Northern Shortgrass Steppe Klipple & Costello (1960), Cover 1940–1942

1952–1953

1984 Ridge

Species

406

Grasses Bouteloua gracilis Buchloë dactyloides Aristida longiseta Sporobolus cryptandrus Agropyron smithii Stipa comata Sitanion hystrix Carex eleocharis Muhlenbergia torreyi Festuca octoflora Forbs Sphaeralcea coccinea Chenopodium album Lepidium densiflorum Salsola iberica Thelesperma spp. Astragalus gracilus

Ashby et. al. (1993), Frequency

Milchunas et al. (1989), Density

H

M

L

U

H

M

L

U

7.41 0.97 0.16 0.03 0.13 0.03 T 0.08 0.09 0.07

6.98 1.38 0.39 0.04 0.09 0.03 0.01 0.06 0.54 0.09

5.90 1.23 0.45 0.04 0.23 0.04 0.01 0.02 0.58 0.10

8.40 0.62 0.18 0.03 0.20 0.01 T 0.01 0.29 0.04

5.48 1.14 0.13 0.02 0.01 0.01 T 0.05 0.07 0.01

5.82 1.37 0.33 0.05 0.02 0.03 0.01 0.04 0.21 0.01

5.34 1.03 0.62 0.02 0.07 0.06 0.01 0.03 0.16 T

6.75 0.54 0.30 0.04 0.07 0.15 0.04 0.07 0.11 T

0.22 0.02

0.20 0.03

0.27 0.03

0.09 0.03

0.02 T

0.02 T

0.06 T

0.05 T

0.04 T 0.04

0.04 T 0.07

0.02 0.02 0.06

0.02 0.02 0.05

T T 0.02

T T 0.01

T 0.01 0.03

T 0.01 T

H

U

1986 Swale

H

U

1569 1409 1921 1491 — — — — 3.1 1.5 3.0 0.2 0.1 0.2 1.1 0.4 1.8 36.0 10.0 193.6 T T T T 0.6 1.9 0.1 0.4 298.0 194.5 474.9 316.9 T T T T 32.3 28.8 27.3 10.4 11.0 T 0.4 0.3 T 0.1

13.4 T 0.3 1.6 T 1.2

9.2 T 0.1 0.6 T 0.3

10.0 T 0.2 1.8 T 0.2

Ridge H

U

1992 Swale

H

U

1340 1174 1748 1272 — — — — 1.6 1.7 1.0 1.1 T T T T 2.5 46.8 8.7 132.8 T 0.5 T 0.1 1.33 2.84 0.09 1.11 337.4 202.6 368.6 358.6 T T T T 9.8 4.1 27.2 1.2 10.3 T 0.5 0.1 T 0.2

16.4 0.1 0.9 1.0 T 1.5

6.4 0.1 0.2 0.2 T T

20.1 0.4 0.2 1.2 0.1 0.2

H

M

L

U

81 14 31 22 1 11 5 72

76 17 32 9 8 0 2 77

57 1 72 37 6 27 10 46

16 7 31 10 29 25 20 59

58 3 20 2 6

54 1 8 0 0

57 4 20 1 16

74 10 27 10 10

Psoralea tenuiflora 0.02 Sophora sericea 0.05 Shrubs Artemisia frigida 0.02 Gutierrezia sarothrae 0.18 Chrysothamnus nauseosus 0.06 Eriogonum effusum 0.22 Cacti Opuntia polyacantha 0.86

0.06 0.14

0.04 0.12

0.03 0.08

T 0.01

T 0.04

0.02 0.02

0.01 T

0.0 0

2.1 0

0.1 0

1.1 0

0.02 0.07 0.44 0.21

0.03 0.06 0.14 0.19

0.19 0.14 0.16 0.11

0.09 0.05 0.07 0.11

0.14 0.03 0.17 0.11

0.15 0.07 0.05 0.12

0.24 0.02 0.07 0.08

0.1 3.8 0.5 6.1

7.6 16.6 2.5 6.1

1.6 0.2 0.4 6.5

0.52

0.74

0.97

0.51

0.40

0.44

0.62

37.4

27.2

14.1

H, heavily grazed; L, lightly grazed; M, moderately grazed; U, ungrazed.

T 0

0.7 0

T 0

1.5 0

19.8 6.7 9.8 10.6

T 0.1 T 0.1

T 0.7 T 0.3

T T T 0.4

0.8 0.1 0.4 1.0

12.4

28.9

29.6

11.4

7.7

1

2

54

46

56

46

76

60

407

408

Ecology of the Shortgrass Steppe

A study of these same treatments in 1962 through 1963, stratifying replicates according to soil type, reported only three species with significant, strong correlations with grazing intensities (Hyder et al., 1966). These and other species varied in the degree and direction of response among soil types. The effects of grazing on species composition were small, and Hyder et al. (1966, 1975) suggested this as a reason to question the usefulness of vegetation-based range condition classification on these grasslands. Hyder et al. (1975) suggested that a management goal for the shortgrass steppe must be to maintain a good stand of B. gracilis, which is thinned after drought but has not been observed to be negatively affected by grazing. More recent assessments of these grazing treatments (Ashby et al., 1993; Hart and Ashby, 1998; Milchunas et al. 1989) show more distinct relationships between grazing intensities and changes in species composition, although conclusions concerning the relatively small magnitude of the overall differences remains similar (Table 16.5). The dominant B. gracilis and other shortgrasses clearly increase with grazing, and less abundant midgrasses generally decrease. This is commonly observed in mixed-grass and tallgrass prairie ecosystems, where grazing can often shift classifications between mixed-grass prairie/shortgrass steppe or tallgrass prairie/mixed-grass prairie (Lauenroth et al., 1994). Convergent, long-term selection forces of semiaridity and grazing have produced plant communities in which large changes in community physiognomy do not occur even after more than 50 years of very large differences in grazing intensity (Milchunas et al., 1988). Early studies of shortgrass steppe plant population dynamics in response to grazing intensity attempted to define general, longer term responses at spatial scales of the whole-pasture management unit. Even though loamy upland sites have only gently rolling topography (Fig. 16.3A,B), grazing by cattle is more intense in swales than on ridgetops, and primary production is often greater in swales (Lauenroth et al., chapter 12, this volume). Differences in the amounts and seasonal patterns of precipitation can favor different species and plant functional types. Interactions among grazing, topography, and short-term weather can be important controls on plant population dynamics (Milchunas et al., 1989). For example, S. coccinea densities were found to decrease from a wet year to a dry year in heavily grazed treatments, but to increase in ungrazed exclosures (Table 16.5). Sensitivity of S. coccinea to grazing during the dry year was in the more favorable swale habitat. Other species display greater sensitivity to grazing during wetyear conditions or in ridgetop communities. Whether a species can be classified as an increaser or a decreaser with grazing depends upon its interactions with other species in a particular community matrix and upon the abiotic conditions. The time of the year that a pasture is grazed can also affect species composition. Repeated heavy grazing during any particular month in the growing season results in approximately three times as many responses in key species as does grazing during any particular month when plants are senescent (Hyder et al., 1975). Heavy grazing in April, May, and June has been found to affect coolseason species most negatively, whereas heavy grazing in September favors coolseason species and best controls annuals. However, weather conditions have been observed to exert greater control over species composition than grazing during

Effects of Grazing on Vegetation 409

any single month of the year at intensities that “could not have been more severe without endangering the lives of the cattle” (Hyder et al., 1975). Grasses are the most abundant life-form of the shortgrass steppe, accounting for approximately 98% of plant density and 80% of aboveground biomass. As a group, grasses generally track increases of the dominant B. gracilis in response to grazing (Fig. 16.8, Table 16.5, and Fig. 16.9). Forbs and shrubs display opposite responses to those of grasses. Both decrease with grazing either because of less tolerance to grazing or because of greater competition with grasses under grazing. Forbs, particularly annuals, are dynamically interactive with respect to

(B)

32

9 7

Succulents (# m-2)

Grasses (thousands m-2)

(A)

LSR 2 4

5 3 1

LSR 2

26 20 14 8 2

G U Swale

G U Ridgetop

W

U

G

D

S

R

LSR 4

Annuals (# m-2)

28

Forbs (# m-2)

30 LSR 4 2

24 18

2

22 16 10

12 4

6

G

U

Swale

G U Swale

G U Swale

G U Ridgetop

Wet year

G U Ridgetop

Dry year

45

LSR 2

35 25 15

U

S

R

W

D

G

U

Swale

G

U

Ridgetop

600 LSR 4 2

400 300 200 100 G

U

Swale

G

U

Dry year

500

0

5

G

Ridgetop

Wet year Cool season species (# m-2)

0

Shrubs (# 100m-2)

D

W

34

G

U

Ridgetop

Wet year

G

U

Swale

G

U

Ridgetop

Dry year

Figure 16.8 Density of functional groups (A) and species or life-forms (B) in relation to long-term grazing treatments (46 and 48 years after initiation) in the northern shortgrass steppe. D, dry year; G, grazed at heavy intensity; R, ridgetop or summit; S, swale or toe slope topographic position; U, ungrazed; W, wet year. Use LSR2 for significance test when crossing any one treatment within the other two treatments and LSR4 when crossing any two treatment categories within a third. A broken x-axis represents a two-way interaction followed by a main effect, or three main effects. (Data from Milchunas et al. [1989].)

410

Ecology of the Shortgrass Steppe (C) Litter basal cover (%)

70

LSR

60

2

50

4

40 30 20 10 0

G

U

G

U

G

Bare ground basal cover (%)

Swale Ridgetop Wet year

G

U

70 LSR 4 2

60 50 40 30 20 10 0

G

U

G

U

G

Swale Ridgetop Wet year

Vegetation basal cover (%)

U

Swale Ridgetop Dry year

U

G

U

Swale Ridgetop Dry year

70 LSR 4 2

60 50 40 30 20 10 0

G

U

G

U

Swale Ridgetop Wet year

G

U

G

U

Swale Ridgetop Dry year

Figure 16.8 (continued) Basal cover (C) in relation to long-term grazing treatments (46 and 48 years after initiation) in the northern shortgrass steppe. D, dry year; G, grazed at heavy intensity; R, ridgetop or summit; S, swale or toe slope topographic position; U, ungrazed; W, wet year. Use LSR2 for significance test when crossing any one treatment within the other two treatments and LSR4 when crossing any two treatment categories within a third. A broken x-axis represents a two-way interaction followed by a main effect, or three main effects. (Data from Milchunas et al. [1989].)

grazing, topography, and short-term, annual wet–dry cycles. Succulents do not respond either to grazing or to short-term fluctuations in weather, but are more abundant on ridgetops than swales (Table 16.5). Cool-season species are a small proportion of the plant community, are more abundant in swales than ridgetops, and can increase or decrease with grazing depending upon topographic position or the species comprising the group. Agropyron smithii is a cool-season grass that is one of the most palatable species in the community and shows very large

Abovegrond net primary production (g m-2 yr-1)

Effects of Grazing on Vegetation 411

130

Shrubs Forbs Cool-season grasses Warm-season grasses

110 90 70 50 30 10 L

M

L

H

M

H

L

M

H

Grazing intensity 1941 - 42 PPT (Oct-Sept) = 312 mm PPT (Oct-April) = 112 mm

1951 - 52 345 mm 84 mm

1991 - 94 387 mm 89 mm

Year and Precipitation Figure 16.9 Aboveground net primary production (measured in grams per square meter per year) of warm- and cool-season grasses, forbs, and shrubs for lightly, moderately, and heavily grazed treatments in the northern shortgrass steppe averaged for years 1941 to 1942, 1951 to 1952, and 1991 to 1994. PPT, precipitation. (Data from Ashby et al. [1993] and Ashby [unpublished data].)

declines with grazing under all situations. The warm-season group is dominated by B. gracilis. The response of annuals to long-term grazing treatments is highly dependent upon topography (Fig. 16.8). Topography influences the interaction effects among soils, water availability, and nutrient budgets (Singh et al., 1998). We have found annuals to have more than four times greater density in heavily grazed than in ungrazed swale communities, but with no significant difference between grazing treatments on ridgetops. Densities of annuals in swale compared with ridgetop communities are very different in wet and dry years. The most abundant annuals are F. octoflora and Plantago patagonica. Both species are unpalatable, very small-stature, short-lived, cool-season species that are possibly favored by grazing as a result of lower litter cover and warmer soil temperatures in grazed than in ungrazed sites during the period of high precipitation in spring. Although annuals are generally more abundant in grazed communities than ungrazed, we found that both native and exotic opportunistic weedy species are more abundant in ungrazed than in heavily grazed communities (Fig. 16.10 [Milchunas et al., 1989, 1990]). Richness of exotic species is also greater in ungrazed compared with grazed communities (see later sections in this chapter).

412

Ecology of the Shortgrass Steppe

Exotic species (# m-2)

85

5

65

Wet year

45

Dry year

7 Number of species 7

25

7

5

8

7

1.0 4

0.8 5

0.6 0.4 0.2

6 2

4

7

2 1

0 Control Grazed Ungrazed Nitrogen Water

Treatment

Grub kill

Water & Nitrogen

Figure 16.10 Density (number of individual plants per square meter) and richness (number of species [indicated by number over bars]) of exotic species in long-term heavily grazed and ungrazed treatments in wet and dry years, and compared with nutrient enrichment and grub kill disturbances in the northern shortgrass steppe. The Control is the control for the nutrient enrichment and the grub kill treatments, and is a short-term exclosure compared with long-term exclosure treatment represented by “ungrazed” treatment. (Data from Milchunas et al. [1989, 1990].)

Differences in community structure between long-term heavily grazed and ungrazed treatments are small (Fig. 16.11), even though some individual species and groups of species respond. However, comparisons between topographic positions indicate that differences between swale and ridgetop communities were greater in ungrazed than in heavily grazed treatments (Fig. 16.3). Segregation of plant communities along the edaphic and microclimatic gradient imposed by landscape topography is expressed in ungrazed treatments, but grazing tends to create a more homogeneous distribution of species across this environmental gradient (Fig. 16.11). Grazers in this particular system also smooth the distributions of soil nutrients and plant biomass across the landscape, but can also create smallscale heterogeneity in species composition by killing long-lived grasses through fecal pat deposition (Peters et al., chapter 6, this volume), by grazing outside of cactus clumps, and by influencing microerosion patterns (see later sections in this chapter). The dual role of grazers in creating both heterogeneity and homogeneity in this system is counter to the common perception of grazers as primarily creating heterogeneity in many other systems. Grazers have often been observed to create heterogeneity by overgrazing particular patches of vegetation (Bakker et al., 1984; Mott, 1987; Vinton et al., 1993), by mediating fire intensity (Hobbs et al., 1991; Vesey-Fitzgerald, 1972), by regulating populations of other consumers

Effects of Grazing on Vegetation 413

1 0.8 0.6 0.4

Similarity (Whittaker index of community association)

0.2

a) Grazed vs. Ungrazed

0 Lowland

Upland

Wet year

Lowland

Upland

Dry year

1 0.8 0.6 0.4 0.2 0

b) Lowlands vs. Uplands Ungrazed

Grazed

Wet year

Grazed

Ungrazed

Dry year

1 0.8 0.6 0.4 0.2 0

c) Wet year vs. Dry year Lowland

Grazed

Upland

Lowland

Upland

Ungrazed

Figure 16.11 Plant community species similarity of heavily grazed compared with ungrazed communities within year and topographic position (A), swales (i.e., Lowland) compared with ridgetops (i.e., Upland) within grazing treatment and year (B), and a wet (1984) compared with a dry (1986) year within grazing treatment and topographic position (C). The long-term treatments were initiated in 1939. Similarity was calculated using the Whittaker (1952) index of community association. A value of one means all species are found in common to both communities and occur in the same proportional abundances in the two communities that are contrasted. A value of zero indicates that there are no species common to both communities being contrasted. Values represent confidence interval range based upon bootstrap method for each replicate–site comparison. (From Milchunas et al. [1989].)

414

Ecology of the Shortgrass Steppe

*1

*1

*1

1.0

*2 R

*2

*1

R S S

0.8

*3

*4

Lightly Grazed vs. Ungrazed Moderately Grazed vs. Ungrazed Heavily Grazed vs. Ungrazed Average All Studies N= 188

0.6

9 8

27 5 26

0.4 0.2 0

1-3

6-8

12-13

45

47

53

30

22 13 13 11 14 11

31

Years of grazing treatment Northern Shortgrass Steppe

N. A me S. Am rica erica Aust ralia Asia Euro pe Afric a

Species similarity (Whittaker index of community association)

that affect vegetation (Noy-Meir, 1988), and by promoting redistribution of soil and soil nutrients (Schlesinger et al., 1990). The relatively small differences in plant community similarities with grazing in the shortgrass steppe can only be appreciated by comparing them with responses in other systems. Species similarity between grazing intensity treatments in the shortgrass steppe are consistently greater than averages for other grasslands in North and South America, Australia, Asia, Europe, and Africa (Fig. 16.12). Changes in species similarities with grazing are most strongly influenced by increasing productivity of the vegetation (Milchunas and Lauenroth,

Southern Shortgrass Steppe

Figure 16.12 Plant community species similarities (Whittaker [1952] index of community association) for northern and southern shortgrass steppe sites, and the years and intensities of grazing treatments compared with average similarities between ungrazed and grazed sites for six other continents. Data for northern shortgrass steppe site are from Klipple and Costello (1960), Milchunas et al. (1989), and Ashby et al. (1993) (all from same treatments at CPER, Colorado); for the southern shortgrass steppe site are from Grant (1971) and Sims et al. (1978), reported in Milchunas and Lauenroth (1993) (Pantex site near Amarillo, Texas) and Vokhiwa (1994) (Comanche Grasslands near Springfield, Colorado); and for the six continents from studies and data reported in Milchunas and Lauenroth (1993). N, number of site comparisons from studies cited in Appendix I of Milchunas and Lauenroth (1993); R, ridgetop community; S, swale community; *1, quadrat cover; *2, density; *3, biomass; *4, line–transect–point cover. See legend for Figure 16.11 for explanation of similarity index.

Effects of Grazing on Vegetation 415

1993; Milchunas et al., 1988), which is consistent with the small changes observed in the low-production shortgrass steppe. Productive plant communities with long evolutionary histories of grazing often display the capacity to shift species composition with changes in grazing pressure (Milchunas and Lauenroth, 1993). The relatively small response of shortgrass steppe communities with a long evolutionary history of grazing may be the result of convergent selection pressures of long-term grazing and semiaridity through evolutionary time. Furthermore, the dominant species increases in abundance with grazing in the shortgrass steppe. Diversity and Dominance The relative and absolute increase of the dominant plant species (B. gracilis) with grazing, as well as the lack of influence of grazing on the abundant Opuntia polyacantha, have a large influence on plant species diversity. In other systems, dominant species most often decline with increasing intensities of grazing (Milchunas and Lauenroth, 1993), and the release in competitive dominance allows for the coexistence of a greater diversity of other species (Milchunas et al., 1988). Concomitant with the increase in B. gracilis with grazing in the shortgrass steppe is a decrease in plant diversity in swales but not on ridgetops (Table 16.6). Within a pasture, swales are grazed more intensively than ridgetops. However, richness is greater in both swales and ridgetops of ungrazed compared with grazed communities. The slight increase in B. gracilis with grazing on upland communities may be countered by the role O. polyacantha plays in maintaining diversity in grazed grassland. This cactus is more abundant in ridgetop than

Table 16.6 Diversity, Richness, Evenness, and Dominance Indices for Long-Term Heavily Grazed and Ungrazed Swales and Ridgetops in Wet and Dry Years for the Northern Shortgrass Steppe Year

Index Value Grazed Swale

Community Attribute Diversity Richness Evenness Dominance

Wet Dry Wet Dry Wet Dry Wet Dry

— X 1.33 1.26 18.7 14.7 0.10 0.09 0.88 0.90

Ungrazed Ridgetop

SD

— X

(0.06) (0.03) (5.0) (8.3) (0.02) (0.01) (0.03) (0.01)

1.31 1.27 18.0 15.0 0.09 0.09 0.89 0.90

Swale

SD

— X

(0.08) (0.05) (2.7) (3.6) (0.02) (0.01) (0.05) (0.03)

1.50 1.46 25.7 18.0 0.13 0.11 0.83 0.84

Ridgetop

SD

— X

SD

(0.10) (0.07) (4.9) (1.7) (0.02) (0.02) (0.04) (0.02)

1.31 1.28 27.0 19.7 0.08 0.08 0.90 0.91

(0.04) (0.03) (1.7) (4.9) (0.01) (0.01) (0.02) (0.02)

Values based upon density data. Diversity calculated as Shannon-Weaver (1949) H⬘. Richness calculated as the total number of species sampled using 48 0.25·m 2 quadrats per each treatment topographic year, and SD based on three replicates each treatment, location, year. Evenness calculated using Pielou (1966), and dominance using Simpson (1949). (Adapted from Milchunas et al. [1989].)

416

Ecology of the Shortgrass Steppe

in swale communities (Table 16.5), and individuals act as microrefugia from grazing (Bayless et al., 1996; Rebollo et al., 2002, 2005). Cattle avoid grazing in the thorny clumps, creating patches of vegetation in heavily and moderately grazed treatments that may be more diverse than outside the clumps (Bayless et al., 1996; Rebollo et al., 2002) (Fig. 16.3C). Cactus protection in long-term moderately grazed treatments has positive effects on some groups of species (Table 16.7). The refuge effect of cactus does not translate into greater plant richness or communities more similar to those that are ungrazed. Most of the effects of cactus are the result of indirect effects of grazing such as litter cover, rather than direct effects of protection from defoliation. Grazing would likely have a greater effect on the plant community if cacti did not create a mosaic of ungrazed patches across the landscape. The unusual response of the shortgrass steppe to grazing (increases in basal cover of total vegetation and of the dominant species, and declines in opportunistic weedy species common in a variety of other disturbed communities) raised questions for shortgrass steppe scientists concerning whether grazing was a disturbance in this system. The predation hypothesis (Paine, 1966, 1971), the intermediate disturbance hypothesis (Fox, 1979; Grime, 1973), and the Huston hypothesis (Huston, 1979, 1985) predict low diversity in undisturbed communities that are dominated by a few superior competitors, highest diversity at intermediate levels of disturbance resulting from suppression of competitive dominants, and low diversity at high levels of disturbance where only a few species are adapted to the harsh conditions. However, in the shortgrass steppe, diversity does not show the predicted bell-shaped (or humpbacked) relationship predicted with increasing intensity of grazing. Diversity (exp. H⬘) calculated from the frequency data of Ashby et al. (1993) (Table 16.5) shows values of 2.5, 2.4, 2.0, and 2.2 for the ungrazed, lightly, moderately, and heavily grazed longterm treatments at the CPER, respectively. Hart (2001a) calculated H⬘ values of 0.71, 0.76, 0.84, and 0.73, respectively, based on biomass at peak standing crop, with values heavily weighted by differences in the large biomass of cactus compared with herbaceous material. However, diversity (exp. H⬘) calculated from frequency data collected the first year of grazing treatment (Klipple and Costello, 1960), before grazing effects could be manifested, are 3.2, 3.3, 3.2, and 3.2, respectively. Similar values of 3.0, 3.2, 3.1, and 3.1, respectively, are obtained from data collected 12 to 13 years after initiating the grazing treatments. In general, the diversity response across grazing intensities appears flat and generally unresponsive through time. It would be reasonable to hypothesize that grazing may not be a disturbance in a system with a long evolutionary history of grazing. However, both the intermediate disturbance and Huston hypotheses were able to predict grazing–diversity relationships in African grasslands, which also have a long evolutionary history of grazing (Milchunas et al., 1988). Evolutionary history alone does not appear to be a good predictor of grazing–diversity relationships. We proposed a model to explain the very different responses in diversity to increasing grazing intensity, based upon interactions with precipitation or productivity, and evolutionary history of grazing (Fig. 16.13)

Table 16.7 Mean Canopy Cover (percent) of Plant Functional Groups inside and outside Cactus Clumps in Ungrazed and Moderately Grazed Treatments since 1939, and Significant Positive and/or Negative Effects of Grazing and Cactus Presence Ungrazed

Grasses Forbs Shrubs Barrel cacti Annuals Perennials Cool season Warm season Cool-season annual grasses Cool-season annual forbs Cool-season perennial grasses Cool-season perennial forbs Warm-season annual forbs Warm-season perennial grasses Warm-season perennial forbs Cool-season shrubs Warm-season shrubs Exotics Weeds Species without asexual reproduction Selected for by cattle Not selected for by cattle Increasers Decreasers Indifferents

Moderately Grazed

In Cactus

Out Cactus

In Cactus

Out Cactus

15.28 4.36 3.31 0.08 0.48 22.54 18.16 4.79 0.00

14.92 5.04 5.91 0.12 0.97 25.03 19.69 6.19 0.00

10.81 4.39 1.35 0.06 0.69 15.92 9.70 6.85 0.006

10.11 3.54 2.38 0.03 1.04 15.02 7.33 8.70 0.05

0.19

0.30

0.11

0.24

11.86

11.15

5.51

3.44

2.99

2.87

2.91

1.78

0.29

0.67

0.57

0.74

3.42

3.77

5.29

6.61

3.42

3.77

5.29

6.61

3.12

5.36

1.15

1.80

0.19

0.55

0.20

0.57

0.18 6.77 18.79

0.34 10.65 21.33

0.14 7.62 14.33

13.02

12.53

9.40 0.16 13.02 9.80

Grazing Effects In Cactus

(–)

– –

Cactus Effects

Out Cactus (–) (–) – – – – (+) (+)

Ungrazed Grazed

– (–) (–) (+)

– –

–

+

–

+

(+) (+)

(–)

–

–

0.10 7.19 11.94

–

– (–) –

(–) – (–)

6.25

4.71

–

–

12.67

9.47

10.64

0.04 14.80 10.97

0.20 4.31 12.12

0.37 3.15 12.47

(+)

–

–

+ –

– (–)

Positive or negative signs within a bracket represent P < .1, and those unbracketed represent P < .05. (Adapted from Rebollo et al. [2002].).

Diversity

Diversity

Ecology of the Shortgrass Steppe

Grazing intensity----> Diversity

Grazing intensity----> Diversity

Evolutionary history of grazing short long

418

Grazing intensity---->

semiarid

Grazing intensity---->

subhumid

Moisture Figure 16.13 Theoretical plant species diversity in relation to grazing intensity for communities in semiarid to subhumid moisture environments and with short and long evolutionary histories of grazing. (From Milchunas et al. [1988].)

Along a precipitation–productivity gradient, competition changes from primarily belowground, for soil water, to increasingly greater competition for light in the aboveground canopy. Short-stature grasslands with large belowground allocation develop in semiarid climates, whereas taller growth forms with greater aboveground allocation develop in grasslands in subhumid environments. Because tall growth forms are generally more susceptible to grazing (Díaz et al., 2007; Milchunas and Lauenroth, 1993), adaptations to grazing and competition for light in a well-developed canopy are divergent selection forces (Milchunas et al., 1988). Short growth forms with relatively greater allocation belowground are adaptations to both herbivory and drought (i.e., convergent selection forces). The evolutionary history of grazing axis represents the time that plants have been exposed to grazing pressures and, thus, the potential for resistance and avoidance mechanisms to have evolved. In subhumid environments, divergent selection forces result in communities with species adapted for either withstanding grazing pressure or competing in a canopy. We expect rapid switches in community composition with changes in grazing pressure (see examples in Milchunas et al., 1988). The shortgrass steppe evolved under convergent selection forces, where removal or imposition of grazing pressure alone is not expected to produce large effects on community composition. Differences between systems with short or long evolutionary histories of grazing within the same precipitation–productivity range result from different capacities to survive and regrow after defoliation. This results in differences in survivorship and the potential for invasion by opportunistic weedy species. There are very few systems such as the shortgrass steppe that have developed under both semiaridity and a long evolutionary history of grazing (Milchunas and Lauenroth, 1993).

Effects of Grazing on Vegetation 419

Invasibility and Comparison with Other Disturbed Communities The greater abundances of weedy species associated with disturbed areas in ungrazed rather than grazed treatments, and the lack of agreement with the intermediate disturbance hypothesis in diversity responses with increasing grazing intensities, raised the question for shortgrass scientists of whether grazing should be considered a disturbance in the shortgrass steppe with its long evolutionary history of grazing. We contrasted communities of both the long-term heavily grazed and the ungrazed treatments with a variety of other disturbances at the CPER (Milchunas et al., 1990). The other disturbances included areas killed by an outbreak of white grubs (Peters et al., chapter 6, this volume) and nutrient enrichment stress plots subjected to large amounts of water, nitrogen, and waterplus-nitrogen. When the ungrazed versus disturbed community comparisons were plotted against the grazed versus disturbed community comparisons, data points fell above the line of equality, in the direction of the ungrazed versus disturbed community comparisons (Fig. 16.14). This indicated that ungrazed communities

Similarity of Ungrazed Lowlands vs Disturbances

1.0 C N CCC N NNC

0.8 0.6 WN WN

0.4

W W

W W W

W

Equality

WN GG G WNWN

0.2 WN 0

0

0.2

0.4

0.6

0.8

1.0

Similarity of Grazed Lowlands vs Disturbances Figure 16.14 Plant community similarity of long-term ungrazed treatments compared with a variety of other disturbed northern shortgrass steppe communities plotted against similarity of longterm heavily grazed treatments compared with the same variety of disturbed communities. Points falling above the line of equality (1:1 relationship) indicate the ungrazed communities are more similar to the disturbed communities than the grazed communities. The disturbed communities were W+N, water plus nitrogen; W, water; and N, nitrogen enrichment applied as stress treatments; G, white grub-killed areas; and C, control versus other control site. r 2 = .88. Data are for swales that are more heavily grazed than ridgetops, calculated from density data, and for 2 years and three replicate sites. See legend for Figure 16.11 for an explanation of the similarity index. (From Milchunas et al. [1990].)

420

Ecology of the Shortgrass Steppe

were more similar to the disturbed communities than were grazed communities. The relationship was more pronounced for the more heavily grazed swale communities than for ridgetops. Although not exactly alike, cattle are a close surrogate for bison grazing, whereby the removal of cattle results in plant communities characteristic of disturbed areas. Although any community can be disturbed by excessive grazing, and grazing by native bison may have occurred in different temporal patterns (Milchunas and Lauenroth, chapter 18, this volume), grazing by domestic livestock appears closer to historical conditions than no grazing at all. Grazing in the shortgrass steppe is similar to the role of fire, flooding, and so forth, in many systems, where these forces are integral, endogenous (sensu Margalef, 1968) components of the system. Establishing “natural” fire regimes for forests, where timing of historical fires is recorded in tree rings, is less subjective than in the case of grasslands. Similarly, there are no good surveys or scientific studies of the numbers or migration patterns of the large herds of bison that inhabited the plains for thousands of years, but Hart (2001b) concluded from journals of early travelers that grazing by bison and other ungulates was heavy and frequent on the Great Plains. Nevertheless, establishing nominal, natural grazing regimes for grasslands of the Great Plains would be difficult and subjective. Responses to grazing discussed thus far suggest that levels of competition and the potential for invasions by exotic species may be greater in ungrazed than grazed communities of the shortgrass steppe. We have examined this indirectly by assessing spatial patterns in root distributions, and directly by seeding plots with five different opportunistic weedy species (Milchunas et al., 1992). We followed seedling establishment and phenological development of individuals in five treatments: (1) long-term heavily grazed, currently grazed during the growing season of study (GG); (2) long-term heavily grazed, currently ungrazed (GU); (3) long-term ungrazed, currently ungrazed (UU); (4) vegetation killed the previous summer with herbicide and structure left intact, currently ungrazed (KU); and (5) vegetation disturbed by blading with a tractor and bare soil hoed, currently ungrazed (DU). This design permitted us to make several comparisons: the direct effects of current-year defoliation compared with the long-term indirect effects of grazing (GG vs. GU), the effects of long-term grazing versus long-term absence of grazing (in the absence of current grazing [i.e., GU vs. UU]); and long-term grazing treatments (GU, UU) compared with initially competition-free plots that were otherwise structurally undisturbed (KU) or disturbed (DU). Kochia scoparia and Salsola iberica established successfully, but emergence of Sisymbrium altissimum, Cirsium arvense, and Lepidium densiflorum was low on all treatments. Of the 500 seeds sown in each quadrat, very few but similar numbers of K. scoparia and S. iberica seedlings established in the GG and GU treatments, whereas very large numbers of seedlings were found in the DU and KU treatments (Fig. 16.15). Nearly four times as many seedlings of S. iberica emerged on the UU treatment than on either GG or GU treatments. Numbers of K. scoparia on UU quadrats were even greater than those emerging on KU, suggesting that the microenvironment created by living plants amplified

Total number of individuals (# yr-1 plot-1)

Effects of Grazing on Vegetation 421 dk

Kochia scoparia

100

Salaola iberica

s

Phenology

a

80

z

60 b

40

bj

j a

k

r r

a ***

GG

i h qx

GU

y ah

qx

UU

y y

t

20 a 0

c

Seedling Juvenile Adult Reproductive

KU

DU

a **

GG

i h qx

GU

qx

UU

KU

DU

Treatment Figure 16.15 Numbers of Kochia scoparia and Salsola iberica individuals reaching seedling, juvenile, adult, and reproductive phenological stages in treatments of GG, long-term heavily grazed, grazed during the experiment; GU, long-term heavily grazed, ungrazed during the experiment; UU, long-term ungrazed, ungrazed during the experiment; KU, vegetation previously killed by herbicide and left intact, ungrazed during experiment, but previously grazed; DU, vegetation previously disturbed by blading and hoeing, ungrazed during experiment. Means within a species and phenological stage not sharing a common letter are significantly different with respect to treatment. Asterisks indicate groups left out of the analysis because of a large number of zeros. Five hundred seeds were sown to each plot, with one species per plot. Species are exotic weeds commonly found along roadsides and other disturbed sites. (From Milchunas et al. [1992].)

germination. Very young seedlings emerging in favorable moisture conditions of early spring probably do not experience competition from established neighbors unless they are in very close proximity. Microenvironmental conditions for germination were more favorable in the UU treatment than the GG and GU treatments. The effects of competition from neighbors probably increases as seed stores are depleted, as root systems of seedlings expand, and as drought periods of summer become more intense. The proportions of the populations that were dead by the following month were high in both the long-term grazed treatments, were intermediate in the long-term ungrazed treatment, and were very low in both competition-free treatments (Table 16.8). By the end of the season, no or very few individuals were found on the GG and GU treatments, and few of these had progressed beyond the seedling stage (Fig. 16.15). A significant number of individuals did survive in the UU treatment, and some of these reproduced. Densities the following spring were 0.4, 0.1, 2.6, 25.9, and 428.0 individuals per quadrat for K. scoparia in the GG, GU, UU, KU, and DU

422

Ecology of the Shortgrass Steppe Table 16.8 Monthly Proportion of Deaths for June Cohort of Kochia scoparia and Salsola iberica Individuals on Plots Sown with 500 Seeds per Plot of One of the Species Monthlya Proportion of Deaths of June Cohort, % of previous month’s population Kochia scoparia Treatment GG GU UU KU DU

Jul 88 a 71 a 43 b 12 c 13 c

Aug c

100 46 a 27 b 7c 13 d

Salsola iberica Sep

Jul

Aug

Sep

— 69 a 40 b 8d 51 bb

84 a 28 b 26 b 6c 12 c

91 a 23 b 23 b 2c 7c

100 c 67 a 32 c 1d 52 bb

a

From previous month to month indicated. High value a result of senescence after flowering rather than premature death. Not included in statistical analyses because of large number of zeros. Treatments: GG, long-term heavily grazed and grazed during the experiment; GU, longterm heavily grazed but ungrazed during the experiment; UU, long-term ungrazed and ungrazed during the experiment; KU, vegetation previously killed by herbicide but left intact and ungrazed during experiment but previously grazed; DU, vegetation previously disturbed by blading and hoeing and ungrazed during experiment. Means within a date and species not sharing a common letter are significantly different with respect to treatment. (Adapted from Milchunas et al. [1992].). b c

treatments, respectively; and 0.0, 0.0, 2.4, 32.6, and 27.4 individuals per quadrat, respectively, for S. iberica. It was not possible to assess clearly the differences in levels of competition between UU and GU treatments because of large differences in seedling emergence. However, it was clear that more favorable microenvironmental conditions for germination of these weedy species existed in long-term ungrazed than in grazed communities, and that the indirect effects of grazing on establishment were more important than the direct effects of current-year defoliation. A greater number of small-mammal disturbances are often observed in ungrazed compared with grazed shortgrass steppe, and this is sometimes thought to be the reason for greater numbers of weeds in ungrazed areas (Ashby et al., 1993; Hart and Ashby, 1998). The very large number of seed-producing individuals in the plots with soil disturbance illustrates that the patches disturbed by small mammals may play a large role in seed availability to surrounding areas, but the emergence in long-term ungrazed, undisturbed plots implicates additional factors in the greater susceptibility to invasion of ungrazed compared with grazed shortgrass steppe. A prolonged cool, recordwet spring occurred in 1995, and these conditions were highly favorable for emergence of Bromus tectorum (annual cheatgrass). Only isolated individuals of this species were previously observed in exclosures, but the large numbers that grew in exclosures in 1995 compared with very few in adjacent grazed grassland substantiate findings from the seed addition experiment that was performed earlier during a year of average precipitation.

Effects of Grazing on Vegetation 423

Distribution of Biomass An important characteristic of grassland plants is the large proportion of biomass in near-surface and belowground organs (Fig. 16.16). Crowns and roots are inaccessible to large herbivores such as cattle, and provide a storage reserve for regrowth after defoliation. In general, the ratio of aboveground to belowground biomass 280

Plant biomass (g m-2)

240

Live leaf Recent dead

Aboveground

200 160 120 80 40 0

Plant biomass (g m-2)

800 600

Surface Old dead Litter

400 200 0 2.6 Crowns Shallow roots Deep roots

Plant biomass (kg m-2)

Belowground 2.0 1.4 0.8 0.2 U G

U

G

U G

U

G

Shortgrass Steppe

Mixedgrass Prairie

northern

northern

southern

southern

U

G

U G

Tallgrass Prairie northern

southern

Figure 16.16 Plant biomass in aboveground (live leaf and recent dead), surface (old dead and litter), and belowground (crown, shallow root, and deep root) categories for ungrazed (U) and grazed (G) treatments at northern and southern shortgrass steppe, northern and southern mixed-grass prairie, and tallgrass prairie sites in the North American Great Plains. (Data from Sims and Singh [1978].)

Ecology of the Shortgrass Steppe

Aboveground : Belowground Biomass (ratio)

424

0.23

Ut Gt

0.20 0.17

Um

Ungrazed: r2 = 0.61

0.14 0.11 0.08

Us 0.05 G s 300

Um Gm

U G ss

Grazed: r2 = 0.90

Gm Um

s = shortgrass steppe m = mixedgrass prairie t = tallgrass prairie

Gm 500

700

900

Precipitation (mm yr-1) Figure 16.17 Aboveground-to-belowground biomass ratios in relation to precipitation (measured in millimeters per year) for ungrazed and grazed treatments in shortgrass steppe, mixedgrass prairie, and tallgrass prairie communities in the North American Great Plains, with trend lines. G, grazed; m, mixedgrass prairie; s, shortgrass steppe; t, tallgrass prairie; U, ungrazed. Aboveground biomass includes live plus recent dead, and belowground is crowns plus roots. (Data from Sims and Singh [1978].)

increases with increasing precipitation (Fig. 16.17). As this ratio increases, defoliation of the canopy has a greater effect on individual plants because a given percentage of aboveground herbivory removes a greater percentage of the total plant biomass. Furthermore, increasing aboveground plant biomass with increasingly productive communities results in increasing levels of competition for light in the canopy. Therefore, alterations in canopy structure (physiognomy) by defoliation will have a greater impact on plant–plant competitive interactions with increasing productivity (Milchunas et al., 1988). We have synthesized grazing studies from around the world (Milchunas and Lauenroth, 1993) and found increasingly greater changes in plant species composition with grazing, and increasingly greater negative effects of grazing on ANPP with increasing aboveground primary production. Comparing results from eight North American grasslands, Sims and Singh (1978) observed that cooler sites had a greater amount of roots in grazed compared with ungrazed treatments, and that warmer sites had no differences or slight reductions of roots in grazed treatments. Several studies have examined root biomass in relation to grazing intensity treatments in the shortgrass steppe. Averaged over seasonal sampling for 3 years, Sims et al. (1978) found no effect of moderate grazing on root biomass at a southern shortgrass steppe site in Texas, and a slightly greater amount in moderately grazed than in ungrazed sites at the CPER in northern Colorado (Fig. 16.16). Samples taken in increments to 80 cm in depth through one growing season indicated no differences in depth distribution of roots at the CPER between ungrazed,

Effects of Grazing on Vegetation 425

Root biomass (kg m-2) 0-10

0.1

0.4 0.7 1.0 1.3 0.1 0.4 0.7 1.0 1.3

Soil profile depth (cm)

10-20 20-40

Lightly grazed

Ungrazed

40-60 60-80 0-10

0.1

10-20 20-40

Moderately grazed

Heavily grazed

40-60

0.4

0.7

1.0

Seasonal average Ungrazed Lightly Moderately Heavily

60-80 Dec

Nov

July

Aug

Figure 16.18 Seasonal distributions of root biomass (measured in kilograms per square meter) by depth in the soil profile (measured in centimeters) for long-term ungrazed, lightly, moderately, and heavily grazed treatments, and seasonal averages for the four grazing treatments, in northern shortgrass steppe communities. (Data from Sims et al. [1971].)

lightly, moderately, and heavily grazed treatments, but somewhat greater seasonal variability at higher grazing intensities (Fig. 16.18 [Sims et al., 1971]). Similarly, Leetham and Milchunas (1985) found no difference in the total biomass or in the depth distribution of roots between the same lightly and heavily grazed treatments at the CPER. Milchunas and Lauenroth (1989) observed slight reductions in 0 to 10-cm root biomass in the heavily grazed compared with ungrazed swales and ridgetops at the CPER, and the same effect of grazing on 10 to 20-cm-deep roots only in swales. For both depths, there were greater differences between topographic positions than there were between grazing treatments. Vokhiwa (1994) found no difference in root biomass or depth distributions between ungrazed and moderately grazed treatments at a southern shortgrass steppe site near Springfield, Colorado. These studies together suggest that there is little to no effect of grazing on root biomass in the shortgrass steppe. In contrast, increases in root biomass with grazing were found at northern mixed-grass prairie sites but not at southern mixed-grass and tallgrass prairie sites (Fig. 16.16). A large effect of grazing on roots was, however, observed in the shortgrass steppe for a distributional characteristic that is seldom examined in root studies. Horizontal spatial distributions of roots and crowns were found to be much more homogeneous in heavily grazed than ungrazed treatments at the CPER (Fig. 16.19). Data from plots that were completely cored in a grid fashion indicated

Grazed

Ungrazed

4 3 2 1 0

Biomass ( g core-1 )

Aboveground 4 3 2 1 0

Crowns 5 4 3 2 1 0

Roots ( 0-10 cm ) 3 2 1 0 0.5

0.5

0.5

0.5

Roots ( 10-20 cm ) Figure 16.19 Horizontal- and vertical-plane spatial distribution of plant biomass (measured in grams per core) in long-term grazed and ungrazed treatments in the northern shortgrass steppe. Data are for one representative plot in a swale topographic position. (From Milchunas and Lauenroth [1989], and see this chapter for means for all plots.)

Effects of Grazing on Vegetation 427

that mean absolute differences between adjacent first-neighbor or second-neighbor cores were sometimes as much as two times greater in ungrazed than in heavily grazed grassland (Milchunas and Lauenroth, 1989). The grazing lawn structure (Fig. 16.3C) of aboveground basal cover and crown biomass under grazed conditions translated to a more uniform exploitation of the soil volume than that found in ungrazed sites. This has potential implications for plant–plant interactions. Microsites favorable for the establishment of invasive opportunistic weedy species may be less available in heavily grazed than in ungrazed shortgrass steppe. Very few weed seedlings established and survived within or directly adjacent to existing plants compared with the numbers found in bare ground or litter (Milchunas et al., 1992). Litter and standing old dead material play important roles in ecosystem function, plant population dynamics, and herbivory. Consumption of plant material, and the return of a portion in the form of urine and feces, reduces the amount of plant material that would otherwise senesce and fall to the surface as litter. Litter affects soil temperature, water dynamics, and the physical structure of the soil surface (Facelli and Pickett. 1991). The large accumulation of litter in ungrazed or unburned highly productive grasslands such as the tallgrass prairie can limit germination and future productivity (Kelting, 1954; Knapp and Seastedt, 1986). The lesser amounts of litter in semiarid regions can still have effects on germination and plant community composition (Milchunas and Lauenroth, 1995; Milchunas et al., 1992), and may play an important role in productivity of grazed versus ungrazed shortgrass steppe (see the next section). Standing dead material from previous-year production not only shades photosynthetically active material, but can also deter grazing by large herbivores as a result of lower quality of forage bites compared with clumps with less dead material. Herbivores often return to areas previously grazed, sometimes leading to patch overgrazing in a heterogeneous mosaic of ungrazed patches in both semiarid (Bridge et al., 1983; Fuls, 1992) and subhumid environments (Bakker et al., 1984; Hunter, 1962; Mott, 1987; Vinton et al., 1993). However, differences between grazed and ungrazed treatments in both litter and standing old dead in the shortgrass steppe are relatively small compared with those in mixed- or tallgrass prairie (Figs. 16.16, 16.20). There is a greater negative effect of grazing on litter biomass with increasing productivity (r2 = 0.69), ranging from–11% in the southern shortgrass steppe to–84% in a North Dakota mixed-grass prairie (Fig. 16.20). Plant Community Productivity Estimates of NPP are subject to a variety of biases and errors (Lauenroth et al., 1986), particularly for belowground components (crown and root) (Milchunas and Lauenroth, 1992, 2001; Singh et al., 1984). Because of serious problems inherent in maxima–minima (peak–trough) methods and the lack of isotope or minirhizotron work, the effects of grazing intensities on roots and crowns will only be addressed here in terms of biomass. Comparisons of traditional methods of estimating ANPP with those obtained using 14C turnover methodology indicate that a single harvest at peak standing crop (in temporarily caged or ungrazed plots) is the best estimate of ANPP in shortgrass steppe communities (Milchunas and Lauenroth, 1992, 2001). The short growing season and dominance by warm-season species

Ecology of the Shortgrass Steppe Difference in litter biomass ([grazed - ungrazed/ungrazed] * 100)

428

-10

-30 r2 = 0.69 -50

-70

-90 170

290 210 250 Aboveground net primary production (g m-2 yr-1)

330

370

Figure 16.20 Difference in litter biomass between grazed and ungrazed treatments (measured as a percentage) in relation to aboveground net primary production (measured in grams per square meters per year) for shortgrass steppe and mixed-grass prairie sites in the Great Plains of North America. (Data from Appendix II in Sims and Singh [1998].)

contribute to a modest underestimate (approximately 16%) of ANPP using this method in the shortgrass steppe compared with greater underestimates that may be obtained in communities with multiple seasonal peaks in the various components of biomass. Estimates of ANPP in grazed systems have the additional problem that the animals are continually consuming plants as they grow. Estimates obtained by clipping within temporary, year-long-placed cages do not account for the potential positive or negative effects of current-year defoliation (McNaughton et al., 1996), but do account for long-term cumulative effects of grazing treatments. We will first examine ANPP in relation to grazing intensities based upon peak standing crop estimates from temporarily caged plots, and then explicitly address the potential for compensatory regrowth resulting from defoliation of the community. Long-Term Effects on Primary Productivity and Seed Production Forage production, excluding cactus and other inedible species, was the most usual method of sampling through the early 1960s. Forage production of the long-term grazing intensity treatments at the CPER spanning a 50-year period averaged 75, 71, 68, and 57 g · m–2 · y–1 for the ungrazed, lightly, moderately, and heavily grazed treatments (n = 15, 89, 97, and 88), respectively (Fig. 16.1). Multiple regression analysis using the long-term forage production data indicated that productivity was most sensitive to amount of cool-season precipitation, followed by amount of warm-season precipitation, then soil fertility, and, last, grazing intensity (Milchunas et al., 1994). Grazing at 20% to 35% annual removal did not alter production. For pastures of average fertility, grazing at a level of 60% removal was

Effects of Grazing on Vegetation 429

predicted to decrease production 3% in wet years and 12% in dry years. The forage available to consumers across the grazing intensity treatments displayed a similar degree of year-to-year variability (standard deviations). This is not consistent with Le Houerou’s (1988) prediction of increasing variability in productivity with increasing grazing intensity. The lack of a grazing treatment-by-precipitation interaction in the regression analyses further suggests that there may not be a decreasing capacity for response to favorable years with increasing grazing intensity. In a study conducted 50 years after initiating the grazing treatments, plots watered to simulate a very wet year compared with unsupplemented plots in a year of average precipitation showed greater differences in ANPP in ungrazed than in lightly or heavily grazed treatments (Varnamkhasti et al., 1995). The longterm ungrazed treatment was more productive than the heavily grazed treatment in the simulated wet year, but the opposite was true under the drier conditions. We conducted this study in swale communities, and mechanically applied defoliation treatments to simulate grazing. Possibly because of these factors and the seasonal distribution of the applied water, the results from this study and the longterm data set analyses do not correspond. However, differences in the response of grazing treatments to large precipitation events versus the usual small events characteristic of drier years suggest a mediating role of litter cover on utilization of precipitation. Higher litter cover in ungrazed treatments may act as an insulator to evaporative loss after large, penetrating events, but may intercept much of the moisture that falls during small precipitation events. Unusual abiotic conditions appear to alter productivity responses of the longterm grazing treatments. This is apparent not only in response to additions of very large precipitation events, but possibly also during recovery from drought. A drought in 1954 resulted in precipitation 62% below average for years in which productivity was monitored on the grazing treatments. Precipitation was +3%,–24%, +30%, and +4% for years 1 through 4 postdrought, respectively. Precipitation use efficiency is a convenient term for comparing postdrought years because it normalizes across different amounts of precipitation by placing productivity on a per-unit-of-precipitation basis. Precipitation use efficiency for 4 years after the drought was reduced in all treatments compared with that for other years of similar precipitation (Fig. 16.21). However, the precipitation use efficiencies were 60% of comparable years in lightly grazed compared with only 41% in moderately and heavily grazed treatments the first year after the drought. Values 4 years after the drought were 75%, 62%, and 52% of those in similar years for the lightly, moderately, and heavily grazed treatments, respectively, and returned to the usual pattern by the fifth year after the drought. Although these data represent only one case, the heavily grazed treatment appeared to be most negatively affected by extreme drought. However, precipitation use efficiencies the second year postdrought were greater in the moderately and heavily treatments than in the lightly grazed treatment, when precipitation was 24% below average. Estimates of ANPP for grazing treatments in the southern shortgrass steppe are limited. Peak standing crops were 59 and 66 g · m–2 in the ungrazed and moderately grazed treatments, respectively, at a Texas site (Lewis, 1971). This compared with 83 and 56 g · m–2 for the respective ungrazed and heavily grazed

430

Ecology of the Shortgrass Steppe

Grazing intensity Precipitation use efficiency (aboveground g m -2 yr -1 mm-1)

Light 2.6

Moderate

Heavy

Mean (yrs similar ppt)

Actual

2.2 1.8 1.4 1.0 0.6 0.2 1

2 3 4

1

2 3 4

1 2 3 4

Years after drought Figure 16.21 Precipitation use efficiency for aboveground plant production (aboveground grams per square meter per year per millimeter precipitation) in relation to long-term light, moderate, and heavy grazing treatments for 1, 2, 3, and 4 years after the severe drought of 1954, and for years of similar precipitation (ppt) to those after the drought. (Data from Milchunas et al. [1994].)

treatments at a northern site. Three years of average standing crop data tend to support differences in the response to grazing of southern versus northern shortgrass sites (Fig. 16.16). Aboveground NPP by functional groups for three periods through the 55-year history of the grazing intensity treatments at the CPER show that weather has a larger effect on ANPP and its botanical composition than grazing. Forbs, shrubs, and cool-season grasses were generally a larger proportion of ANPP, and total ANPP was greater in all grazing treatments in the early 1940s than in the 1950s or 1990s (Fig. 16.9) (Ashby, unpublished data; Ashby et al., 1993; Hart and Ashby, 1998). Cool-season precipitation was lower during the latter two periods, although the total annual amount was greater. Cool-season precipitation has a greater effect on productivity than warm-season precipitation, and particularly on the productivity of forbs and cool-season grasses, both of which are sensitive to grazing intensity. We found greater differences in ANPP between ungrazed and heavily grazed treatments in swales than in upland communities (Milchunas et al., 1992). Grazing also has an effect on the spatial distribution of B. gracilis seed production. Soil texture has a significant influence on seed production of B. gracilis in ungrazed shortgrass steppe (greatest on sandy soils), but not in grazed sites (Coffin and Lauenroth, 1992). Ungrazed sites have been found to have twice as much viable seed production per flowering culm, and greater seed production per area as well. In contrast, Rebollo et al. (2002) found greater overall seedhead production in moderately grazed compared with ungrazed treatments. This was

Effects of Grazing on Vegetation 431

primarily because of the increase in seedhead production of B. gracilis inside cactus clumps in the moderately grazed treatment. Grazing also has positive effects on the number of inflorescents per unit of cover for shrubs and cool-season annual grasses and forbs, and negative effects on cool-season perennial grasses. Compensatory Regrowth: Interactions with Grazing History McNaughton (1979) found greater productivity where plants were defoliated during the current-year growing season than where they were not. Williamson et al. (1989) grazed grasshoppers for short periods, caged at different densities, in level upland communities that had not been grazed by cattle for 10 to 11 years. When grazing season precipitation was higher than average, current-year defoliation by grasshoppers increased ANPP in only a few cases. However, when grazing season precipitation was lower than average, increases in ANPP were found across a range of grasshopper grazing intensities. These authors concluded that compensatory regrowth was most likely to occur during short-term recovery from dry periods. In a study not designed to assess compensatory regrowth, Milchunas et al. (1992) observed that end-of-season biomass was the same in long-term heavily grazed treatments regardless of whether they were grazed during that particular year. This suggested that compensatory regrowth in grazed areas was as great as the amount that had been removed by the cattle. The study by Williamson et al. (1989) and the observation by Milchunas et al. (1992) led to a more detailed study of interactions among grazing history, precipitation, and current-year-defoliation in compensatory regrowth responses. There are two methods to assess the effects of current-year grazing by large herbivores on ANPP (McNaughton et al., 1996) that mimic the actual pattern and frequency of grazing by the animals: (1) moving cages through the growing season and clipping caged and grazed plots, and (2) observing the pattern of bites taken from grazed reference plots and then clipping caged plots in the same manner. The latter method was developed and applied in long-term heavily and lightly grazed areas to which either supplemental water was added to simulate a very wet year or no water was added in a year of average precipitation (Varnamkhasti et al., 1995). Water treatment had a greater overall effect on ANPP than did either longterm grazing history or current-year defoliation, although all three factors interactively determined ANPP (Fig. 16.22). Slightly higher ANPP and precipitation use efficiency of the unwatered, long-term lightly grazed treatment with defoliation resulted in different relationships between grazing and watering treatments. These data and the results of Williamson et al. (1989) suggest that defoliation may increase ANPP during periods of water stress, but that some of the potential for regrowth may diminish with heavy grazing. Other studies have reported conservation of soil water or improved plant–water relations in response to grazing or defoliation (Archer and Detling, 1986; Baker and Hunt, 1961; Coughenhour et al., 1990; Day and Detling, 1994; Eck et al., 1975; Hodgkinson, 1976; McNaughton et al., 1983; Reed and Dwyer, 1971; Simoes and Baruch, 1991), although only a few observed increased ANPP. Young leaves can have a lower transpiration loss per unit of growth than older leaves removed by defoliation.

432

Ecology of the Shortgrass Steppe

Watering Control 130

Addition

ungrazed

heavily

lightly

110 90 70

Net primary production (aboveground g m -2 y -1)

50 30 10 Nondefoliated Defoliated

Nondefoliated Defoliated

Clipping 32

130

32

32

90 70

32

32

110 32

50 30 10 U

L

U

H

L

H

Grazing Figure 16.22 Aboveground net primary production (measured in grams per square meter per year) in long-term lightly and heavily grazed, clipped or not clipped, and control or watered treatments (A); and ungrazed, and lightly and heavily grazed treatments (with clipping), and control or watered treatments (B). Clipping treatments (D, defoliated; N, nondefoliated) simulated natural patterns and intensities observed in adjacent uncaged reference plots. Watering treatments were an unwatered control during a year of average precipitation and a watered treatment equivalent to very wet years in the northern shortgrass steppe. Confidence intervals (Tukey’s Honest Significant Difference) are for differences between means of any one type of treatment, holding the level of the other treatments constant. H, heavily grazed; L, lightly grazed; U, ungrazed. (From Varnamkhasti et al. [1995].)

Small effects of defoliation on compensatory regrowth were observed in the previously mentioned studies at the CPER. At the same research site but in more productive sandy soils, very large differences between mid growing seasondefoliated and -nondefoliated treatments were observed in two separate studies. In a study of the effects of elevated levels of atmospheric CO2, Morgan et al. (2001)

Effects of Grazing on Vegetation 433

found an ANPP of 78 g · m–2 · y–1 on nondefoliated compared with 134 g · m–2 · y–1 on defoliated treatments, and no interaction with CO2 treatment. Milchunas et al. (2005) observed 38 versus 50 g · m–2 · y–1 of digestible forage production for nondefoliated and defoliated treatments averaged over 4 years and all CO2 treatments, and no interaction with CO2 treatment. This site had been lightly to moderately grazed in the past. Differences between the responses observed in this study and the smaller compensatory responses discussed earlier could be the result of the greater ANPP at the CO2 study site or because there was a single defoliation during the season of entire (but small and alternating) quadrats. Similar conditions of greater ANPP than at sites reported earlier, previously long-term moderately grazed, and clipping of entire quadrats once at mid growing season were used in a study of ultraviolet radiation and defoliation effects on ANPP, and a very wet and a very dry year were simulated (Milchunas et al., 2004). Aboveground NPP of nondefoliated treatments averaged 170 g · m–2 · y–1 compared with 290 g · m–2 · y–1 for defoliated treatments in the very wet year, and there was no defoliation effect for the dry year treatment when ANPP averaged 50 g · m–2 · y–1. The interaction of current-year defoliation with level of precipitation differed in this study compared with the studies reported earlier. This may be the result of the severe drought imposed in the ultraviolet study or any of the other differences among the studies, but suggests that the potential for compensatory regrowth after current-year defoliation may be affected by many factors and is not easily defined. Forage Quality and Consumption Conclusions concerning forage quality in relation to grazing intensity treatments in the northern shortgrass steppe differ depending upon the study and/or the methods of collecting samples. Uresk and Sims (1975) found no difference in nitrogen concentrations of live or dead B. gracilis hand-plucked from long-term ungrazed, lightly, moderately, and heavily grazed treatments for 11 sample dates in each of 2 years. Similarly, Milchunas (unpublished data) observed no differences in nitrogen concentrations of B. gracilis, S. coccinea, or G. sarothrea (grass, forb, and shrub, respectively) hand-plucked from moderately grazed and 13- to 17-year-old exclosures, averaged over 5 years and 17 sampling dates. However, differences between the grazed and the ungrazed treatment were sometimes positive and sometimes negative, depending upon the sample date and year. In a more controlled experiment, Milchunas et al. (1995) assessed nitrogen content and digestibility of vegetation from entire plots that were subjected to defoliation and/or water supplementation in long-term ungrazed, lightly, and heavily grazed treatments. Effects of grazing on nitrogen concentration and on percent in vitro digestible dry matter (Fig. 16.23) were similar to those observed for productivities (Fig. 16.22), with the effects being more pronounced on quality than on quantity of vegetation. Multiplying the biomass productivities by the concentrations of nitrogen and digestible dry matter to estimate nitrogen and digestible forage yield accentuated differences between the treatments. We observed large increases in nitrogen yield with water addition in the ungrazed treatment, but not in the lightly grazed treatment (Fig. 16.24B), probably because of the already large stimulation in nitrogen yield with defoliation (Milchunas et al., 1995). Similar

Watering Control

Addition

Digestibility (% invitro dry matter)

70

2

2

3

2

3

2

3

3 2 3

65

2 3

60

55

50 2 3

Nitrogen concentration (%)

2.0 1.8

2 3

2 3 2 3

2 3 2 3

1.6 1.4 1.2 1.0

U

L

H

Grazing

U

L

H

Figure 16.23 Nitrogen and in vitro digestible dry matter concentrations (measured as a percentage) of vegetation in ungrazed, lightly, and heavily grazed treatments, with or without water addition. The grazed treatments were clipped to simulate natural patterns and intensities observed in adjacent uncaged reference plots. Watering treatments were an unwatered control during a year of average precipitation and a watered treatment equivalent to very wet years in the northern shortgrass steppe. Confidence intervals (Tukey’s Honest Significant Difference) are for differences between means of any one type of treatment holding the level of the other treatments constant. H, heavily grazed; L, lightly grazed; U, ungrazed. (From Milchunas et al. [1995].)

Digestible forage production (invitro digestible dry matter, g m-2 yr-1)

Watering

90

23 23

80

23

23

70 23

60 50 23

40 30 23

2.6

Nitrogen Yield (g m-2 yr-1)

Addition

Control

23

2.0 1.4

23

2 3

23 23

0.8 0.2 U

L

H

U

L

H

Grazing

Figure 16.24 Nitrogen and in vitro digestible dry matter yields (measured in grams per square meter per year) of vegetation in ungrazed, lightly, and heavily grazed treatments, with or without water addition. The grazed treatments were clipped to simulate natural patterns and intensities observed in adjacent uncaged reference plots. Watering treatments were an unwatered control during a year of average precipitation and a watered treatment equivalent to very wet years in the northern shortgrass steppe. Confidence intervals (Tukey’s Honest Significant Difference) are for differences between means of any one type of treatment holding the level of the other treatments constant. H, heavily grazed; L, lightly grazed; U, ungrazed. (From Milchunas et al. [1995].)

436

Ecology of the Shortgrass Steppe

relationships between the ungrazed, lightly, and heavily grazed treatments were found for digestible forage production (Fig. 16.24A), but the effects of defoliation on digestible forage production were less than for nitrogen yield. Because of the importance of forage quality to consumers, ANPP may not be the most important indicator for assessing the effects of grazing. For instance, cattle weight gain on the heavily grazed treatment is more negatively affected than ANPP (Hart, chapter 4, this volume), and pronghorn are able to maintain similar crude protein levels in their diets throughout the year when grazing the lightly or heavily grazed pastures, even though the quantity of forage was lower on the heavily grazed treatment (Schwartz et al., 1977). Increased forage quality with currentyear defoliation may partially compensate for declining productivity and biomass availability, but the capacity for this to occur is diminished with long-term heavy grazing intensities. The lightly grazed pastures provided more stable year-to-year yields of forage quality than either the ungrazed or heavily grazed treatments, and provided greater quality yields in the year of average precipitation. Comparison of Response with That of Other Ecosystems We found that the average reduction in ANPP with grazing for all shortgrass steppe studies and treatments (Milchunas and Lauenroth, 1993) was–10%, at an average grazing intensity of 44% (n = 10). For similar grazing intensities, northern and southern mixed-grass prairie sites display average reductions in ANPP of–22% and–23%, respectively. This agrees with the statistical model derived from a worldwide data set that shows increasingly greater negative effects of grazing on ANPP with increasing productivity of a site. The model by Milchunas and Lauenroth (1993) suggests that declines in ANPP with grazing are more than twice as sensitive to increasing site productivity than they are to increasing intensities of grazing. An anomaly is the potential for grazing to increase productivity in very productive grasslands where litter buildup can inhibit germination or emergence of vegetation. For instance, tallgrass sites increased 18% with grazing of unburned sites (average grazing intensity, 36%; n = 4). Similar to the relatively small ANPP response, grazing may be expected to have relatively less impact on forage quality in the shortgrass steppe compared with other more productive grasslands (Milchunas et al., 1995). We base this conclusion on the relatively minor impact of grazing on community physiognomy and on species composition in this system. Increases in less palatable species with grazing can decrease overall forage quality (Cook and Harris, 1968; Pieper et al., 1959; Westoby, 1985, 1986), but this, as well as high stem-to-leaf and deadto-live ratios can more readily develop in communities with taller growth forms (Coppock et al., 1983; Fryxell, 1991; McNaughton, 1984).

Summary and Overview Plant species composition in the shortgrass steppe is very resistant to grazing, more so than other grasslands (Fig. 16.25). Changes that take place in response

100 11

(Grazed vs Ungrazed %)

Plant Community Species Dissimilarity

Effects of Grazing on Vegetation 437 14

36

26

32

45

Arid/ Short History

Subhumid/ Short History

58

80

60

40

20

0

Fire Grazing

SGS

Grassland

Southwest US

Arid/ Long History

Subhumid/ Long History

The Rest of the World

Figure 16.25 Plant community species dissimilarity of burned versus unburned and grazed versus ungrazed treatments in the northern shortgrass steppe compared with species dissimilarities of grazed versus ungrazed grasslands in the adjacent southwestern United States (with a short evolutionary of grazing and similar productivity) and grazed versus ungrazed rangelands around the world. (Fire data are from Milchunas [unpublished data]; shortgrass steppe grazing data are from Klipple and Costello [1960], Grant [1971], Milchunas et al. [1989], and Ashby et al. [1993]; data for the southwestern United States is from Milchunas [2006], and data from around the world from 276 grazed versus ungrazed comparisons are from Milchunas and Lauenroth [1993].) Dissimilarity is converse of similarity, as calculated and described in the legend for Figure 16.11.

to long-term grazing indicate that ungrazed, rather than grazed, plant communities are more likely to be invaded by exotic and native weedy species. Ungrazed communities, then, are more representative of disturbed plant communities than are grazed communities. The mechanism for this appears to be the increased spread (basal cover) by native prostrate growth forms in grazed areas, resulting in a more uniform exploitation of the soil volume and reductions in safe sites for weedy colonization. This resistance of the shortgrass plant community to grazing pressure includes components of both avoidance and tolerance. A long evolutionary history of grazing by large native herbivores (bison) that are ecologically similar to cattle produced plant species tolerant of defoliation. The semiarid environment produced a plant community with a large proportion of biomass belowground that is inaccessible to large mammalian herbivores, and one with a sparse canopy in which the potential for herbivore-mediated alteration of competition for light is minimal. The long history of grazing and semiaridity act as convergent selection forces providing resistance to current-day grazing pressure (Milchunas et al., 1988).

438

Ecology of the Shortgrass Steppe

These attributes are exemplified in the dominant species B. gracilis. Because a large proportion of B. gracilis biomass is in its roots it is not severely affected by grazing by large herbivores. Furthermore, its dominance in the community may give it some protection from grazing. Although a palatable and nutritious plant, it is consumed less than its proportional abundance in the community, as grazers seem to prefer a mixed diet, possibly to balance their nutritional needs. Grazing of livestock on public lands has been and continues to be a controversial issue. Proponents of the removal of livestock from public lands often fail to distinguish between systems with long versus short evolutionary histories of grazing and the potential for this to result in very different outcomes in responses to grazing (Milchunas, 2006). Similarly, administration of grazing fees and conservation programs by federal agencies is generally applied uniformly across all rangelands. A conclusion of the global analyses of grazing impacts by Milchunas and Lauenroth (1993) was that, “within levels not considered abusive ‘overgrazing,’ the geographical location where grazing occurs may be more important than how many animals are grazed or how intensively an area is grazed” (p. 327). This obviously does not mean that good local grazing management is not important, but that the larger regional scale of grazing management, particularly with respect to the geography of evolutionary history, is important as well. When viewed at a regional scale, where we graze is more important than how we graze. Traditional range management focuses on individual ranch-scale practices, and much progress has been made during the past century. Differential application of policy at larger scales based upon what we know of system response to grazing is more difficult, but offers potential for balancing environmental and production needs. It is interesting to note that the southwestern United States is considered to be a hotspot of species endangerment (Flather et al., 1994). This is an area where grazing is currently an important land use of native ecosystems that have only short histories of grazing by large herbivores. Similar to fire, grazing can be a devastating force in some systems but an integral part of others. Milchunas (2006) and Milchunas et al. (1998) suggest that Margalef’s (1968) terminology of endogenous and exogenous disturbances may provide a useful conceptual framework for sociopolitical issues concerning utilization of rangelands by domestic livestock. Although any system can be overgrazed, moderate grazing of the shortgrass steppe by domestic livestock more closely resembles the endogenous disturbance of bison grazing than does protection from grazing in this particular system. Endogenous disturbances are a stabilizing force in a system in which the flora and fauna have evolved in concert with the disturbance.

Acknowledgments Rod Heitschmidt, Paul Stapp, and Dick Hart provided reviews of early versions of the manuscript. We thank LTER site managers Mark Lindquist, Don Hazlett, and Ray Souther, and the many students on the LTER field crews, for a constant source of help in field sampling, and Judy Hendryx for support and coordination in the laboratory.

Effects of Grazing on Vegetation 439 References Archer, S., and J. K. Detling. 1986. Evaluation of potential herbivore mediation of plant water status in a North American mixed-grass prairie. Oikos 47:287–291. Ashby, M. M., R. H. Hart, and J. R. Forewood. 1993. Plant community and cattle responses to fi fty years of grazing on shortgrass steppe. USDA–ARS, Rangeland Resources Research Unit, RRRU-1, Fort Collins, Colo. Axelrod, D. I. 1985. Rise of the grassland biome, central North America. Botanical Review 51:163–201. Baker, J. N., and O. J. Hunt. 1961. Effects of clipping treatments and clonal differences on water requirement of grasses. Journal of Range Management 14:216–219. Bakker, J. P., J. de Leeuw, and S. E. van Wieren. 1984. Micro-patterns in grassland vegetation created and sustained by sheep grazing. Vegetatio 55:153–161. Bayless, M., W. K. Lauenroth, and I. C. Burke. 1996. Refuge effect of the Opuntia polyacantha (plains prickly pear) on grazed areas of the shortgrass steppe. Bulletin of the Ecological Society of America. Bement, R. E. 1993. Colorado rangelands: A land manager’s historical perspective. Rangelands 15:208–210. Bridge, B. J., J. J. Mott, and R. J. Hartigan. 1983. The formation of degraded areas in the dry savanna woodlands of northern Australia. Australian Journal of Soil Research 21:91–104. Briske, D. D., and V. J. Anderson. 1992. Competitive ability of the bunchgrass Schizachyrium scoparium as affected by grazing history and defoliation. Vegetatio 103:41–49. Brougham, R. W., and W. Harris. 1967. Rapidity and extent of changes in genotypic structure induced by grazing in rye grass populations. New Zealand Journal of Agricultural Research 10:56–65. Coffin, D. P., and W. K. Lauenroth. 1992. Spatial variability in seed production of the perennial grass Bouteloua gracilis (H.B.K.) LAG. ex Griffiths. American Journal of Botany 79:347–353. Cook, C. W., R. D. Child, and L. L. Larson. 1977. Digestible protein in range forages as an index to nutrient content and animal response. Colorado State University, Range Science Department, science series no. 29, Fort Collins, Colo. Cook, C. W., and L. E. Harris. 1968. Nutritive value of seasonal ranges. Utah State University, Agricultural Experiment Station Bulletin 472, Logan, Utah. Cook, C. W., and E. F. Redente. 1993. Development of the ranching industry in Colorado. Rangelands 15:204–207. Coppock, D. L., J. K. Detling, J. E. Ellis, and M. I. Dyer. 1983. Plant–herbivore interactions in a North American mixed-grass prairie. I. Effects of black-tailed prairie dogs on intraseasonal aboveground plant biomass and nutrient dynamics and plant species diversity. Oecologia 56:1–9. Coughenour, M. B., J. K. Detling, I. E Bamberg, and M. M. Mugambi. 1990. Production and nitrogen responses of the African dwarf shrub Indigofera spinosa to defoliation and water limitation. Oecologia 83:546–552. Daubenmire, R. F. 1985. The western limits of the range of the American bison. Ecology 66:622–624. Day, T. A., and J. K. Detling. 1994. Water relations of Agropyron smithii and Bouteloua gracilis and community evapotransporation following long-term grazing by prairie dogs. American Midlands Naturalist 132:381–392.

440

Ecology of the Shortgrass Steppe

Detling, J. K. 1989. Grasslands and savannas: Regulation of energy flow and nutrient cycling by herbivores, pp. 131–148. In: L. R. Pomeroy and J. J. Alberts (eds.), Concepts of ecosystem ecology: A comparative view. Ecological studies 67. SpringerVerlag, New York. Detling, J. K., and E. L. Painter. 1983. Defoliation responses of western wheatgrass populations with diverse histories of prairie dog grazing. Oecologia 57:65–71. Díaz., S., S. Lavorel, S. McIntyre, V. Falczuk, F. Casanoves, D. G. Milchunas, C. Skarpe, G. Rusch, M. Sternberg, I. Noy-Meir , J. Landsberg, Z. Wei, H. Clark, and B. D. Campbell. 2007. Plant trait responses to grazing: A global synthesis. Global Change Biology 13:313–341. Eck, H. V., W. G. McCully, and J. Stubbendieck. 1975. Response of shortgrass plains vegetation to clipping, precipitation, and soil water. Journal of Range Management 28:194–197. Facelli, J. M., and S. T. A. Pickett. 1991. Plant litter: Its dynamics and effects on plant community structure. Botanical Review 57:1–32. Flather, C. H., L. A. Joyce, and C. A. Bloomgarden. 1994. Species endangerment patterns in the United States. United States Department of Agriculture Forest Service, general technical report RM-241. Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colo. Fox, J. F. 1979. Intermediate-disturbance hypothesis. Science 204:1344–1345. Fryxell, J. M. 1991. Forage quality and aggregation by large herbivores. The American Naturalist 138:478–498. Fuls, E. R. 1992. Ecosystem modification created by patch-overgrazing in semi-arid grassland. Journal of Arid Environments 23:59–69. Grant, W. E. 1971. Comparisons of aboveground plant biomass on ungrazed pastures vs. pastures grazed by large herbivores, 1970 season. U.S. International Biological Program, Grassland Biome technical report no. 131. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. Grime, J. P. 1973. Control of species density in herbaceous vegetation. Journal of Environmental Management 1:151–167. Hansen, R. M., and I. K. Gold. 1977. Blacktail prairie dogs, desert cottontails and cattle trophic relations on shortgrass range. Journal of Range Management 30:210–214. Harper, J. L. 1969. The role of predation in vegetational diversity, pp. 48–62. In: G. M. Woodwell and H. H. Smith (eds.) Diversity and stability in ecological systems. Brookhaven Symposia in Biology no. 22. National Technical Information Service, Springfield, Va. Hart, R. H. 2001a. Plant biodiversity on shortgrass steppe after 55 years of zero, light, moderate, or heavy cattle grazing. Plant Ecology 155(1):111–118. Hart, R. H. 2001b. Where the buffalo roamed: Or did they? Great Plains Research 11:83–102. Hart, R. H., and M. M. Ashby. 1998. Grazing intensities, vegetation, and heifer gains: 55 years on shortgrass. Journal of Range Management 51:392–398. Heitschmidt, R. K., S. L. Dowhower, R. A. Gordon, and D. L. Price. 1985. Response of vegetation to livestock grazing at the Texas Experimental Ranch. Texas Agriculture Experiment Station, B-1515, College Station, Texas. Hobbs, N. T., D. S. Schimel, C. E. Owensby, and D. J. Ojima. 1991. Fire and grazing in the tallgrass prairie: Contingent effects on nitrogen budgets. Ecology 72:1374–1382.

Effects of Grazing on Vegetation 441 Hodgkinson, K. C. 1976. The effects of frequency and extent of defoliation, summer irrigation, and fertilizer on the production and survival of the grass Danthonia caespitosa Gaud. Australian Journal of Agricultural Research 27:755–767. Hornaday, W. T. 1889. The extermination of the American bison with a sketch of its discovery and life history. Smithsonian report no. 1887, Smithsonian Institution, Washington, D.C. Hunter, R. F. 1962. Hill sheep and their pasture: A study of sheep grazing in southeast Scotland. Journal of Ecology 50:651–680. Huston, M. 1979. A general hypothesis of species diversity. The American Naturalist 113:81–101. Huston, M. A. 1985. Patterns of species diversity on coral reefs. Annual Review of Ecology Systematics 16:149–177. Hyder, D. N., R. E. Bement, E. E. Remmenga, and D. F. Hervey. 1975. Ecological responses of native plants and guidelines for management of shortgrass range. United States Department of Agriculture–Agricultural Research Service, technical bulletin no. 1503. U.S. Government Printing Office, Washington, D.C. Hyder, D. N., R. E. Bement, E. E. Remmenga, and C. Terwillager, Jr. 1966. Vegetation– soils and vegetation–grazing relations from frequency data. Journal of Range Management 19:11–17. Jaramillo, V. J., and J. K. Detling. 1988. Grazing history, defoliation, and competition: Effects on shortgrass population and nitrogen accumulation. Ecology 69:1599–1608. Johnson, C. W. 1951. Protein as a factor in the distribution of the American bison. Geography Review 41:330–331. Joyce, L. A. 1989. An analysis of the range forage situation in the United States: 1989–2040. United States Department of Agriculture, general technical report RM-180. Forest Service, Rocky Mountain Forest and Range Experiment Station. Fort Collins, Colo. Kelting, R. W. 1954. Effects of moderate grazing on the composition and plant production of a native tallgrass prairie in central Oklahoma. Ecology 35:200–207. Kemp, W. B. 1937. Natural selection within plant species as exemplified in a permanent pasture. Journal of Heredity 28:329–333. Kingston, C. S. 1932. Buffalo in the Pacific Northwest. Washington Historical Quarterly 23:163–172. Klipple, G. E., and D. F. Costello. 1960. Vegetation and cattle responses to different intensities of grazing on short-grass ranges on the central Great Plains. United States Department Agriculture technical bulletin no. 1216. U.S. Government Printing Offfice, Washington D.C. Knapp, A. K., and T. R. Seastedt. 1986. Detritus accumulation limits productivity of tallgrass prairie. BioScience 36:662–668. Larson, F. 1940. The role of the bison in maintaining the short grass plains. Ecology 21:113–121. Lauenroth, W. K., H. W. Hunt, D. M. Swift, and J. S. Singh. 1986. Estimating aboveground net primary production in grasslands: A simulation approach. Ecological Modeling 33:297–314. Lauenroth, W. K., and D. G. Milchunas. 1991. The shortgrass steppe, pp. 183–226. In: R. T. Coupland (ed.), Natural grasslands, introduction and western hemisphere. Ecosystems of the world 8A. Elsevier, Amsterdam.

442

Ecology of the Shortgrass Steppe

Lauenroth, W. K., D. G. Milchunas, J. L. Dodd, R. H. Hart, R. K. Heitschmidt, and L. R. Rittenhouse. 1994. Grazing in the Great Plains of the United States, pp. 69–100. In: M. Vavra, W. A. Laycock, and R. D. Pieper (eds.), Ecological implications of livestock herbivory in the West. American Institute of Biological Sciences Symposium, San Antonio, Texas, 1991. Society for Range Management, Denver, Colo. Leetham, J. W., and D. G. Milchunas. 1985. The composition and distribution of soil microarthropods in the shortgrass steppe in relation to the soil water, root biomass, and grazing by cattle. Pedobiologia 28:311–325. Le Houerou, H. N. 1988. Relationship between the variability of primary production and the variability of annual precipitation in world arid lands. Journal of Arid Environments 15:1–18. Leopold, E. B., and M. F. Denton. 1987. Comparative age of grassland and steppe east and west of the northern Rocky Mountains. Annals of the Missouri Botanical Gardens 74:841–867. Lewis, J. K. 1971. The grassland biome: A synthesis of structure and function, 1971, pp. 317–387. In: N. R. French (ed.), Preliminary analysis of structure and function in grasslands. Range Science Department science series no. 10. Colorado State University, Fort Collins, Colo. Liang, Y. M., D. L. Hazlett, and W. K. Lauenroth. 1989. Biomass dynamics and water use efficiencies of five plant communities in the shortgrass steppe. Oecologia 80:148–153. Mack, R. N., and J. N. Thompson. 1982. Evolution in steppe with few large, hooved mammals. The American Naturalist 119:757–773. Margalef, R. 1968. Perspectives in ecological theory. University of Chicago Press, Chicago, Ill. McDonald, J. N. 1981. North American Bison: Their classification and evolution. University of California Press, Berkeley, Calif. McHugh, T. 1972. The time of the buffalo. A. A. Knopf, New York. McNaughton, S. J. 1979. Grazing as an optimization process: Grass–ungulate relationships in the Serengeti. The American Naturalist 113:691–703. McNaughton, S. J. 1984. Grazing lawns: Animals in herds, plant form, and coevolution. The American Naturalist 124:863–886. McNaughton, S. J. 1985. Ecology of a grazing ecosystem: The Serengeti. Ecological Monographs 55:259–294. McNaughton, S. J., D. G. Milchunas, and D. A. Frank. 1996. How can productivity be measured in grazing ecosystems? Ecology 77:974–977. McNaughton, S. J., L. L. Wallace, and M. B. Coughenour. 1983. Plant adaptation in an ecosystem context: Effects of defoliation, nitrogen, and water on growth of an African C4 sedge. Ecology 64:307–318. Mengel, R. M. 1970. The North American Central Plains as an isolating agent in bird speciation, pp. 279–340. In: W. Dort and J. K. Jones (eds.), Pleistocene and recent environments of the central Great Plains. University Kansas Department Geology special publication no. 3. University of Kansas Press, Lawrence, Kans. Milchunas, D. G. 2006. Responses of plant communities to grazing in the southwestern United States. General technical report RMRS–GTR-169. USDA Forest Service, Rocky Mountain Research Station, Fort Collins, Colo.

Effects of Grazing on Vegetation 443 Milchunas D. G., J. R. Forwood, and W. K. Lauenroth. 1994. Forage production across fifty years of grazing intensity treatments in shortgrass steppe. Journal of Range Management 47:133–139. Milchunas, D. G., J. Y. King, A. R. Mosier, J. C. Moore, J. A. Morgan, M. H. Quirk, and J. R. Slusser. 2004. UV radiation effects on plant growth and forage quality in a shortgrass steppe ecosystem. Phytochemical Phytobiology 79:404–410. Milchunas, D. G., and W. K. Lauenroth. 1989. Three-dimensional distribution of plant biomass in relation to grazing and topography in the shortgrass steppe. Oikos 55:82–86. Milchunas, D. G., and W. K. Lauenroth. 1992. Carbon dynamics and estimates of primary production by harvest, C14 dilution, C14 turnover. Ecology 73:593–607. Milchunas, D. G., and W. K. Lauenroth. 1993. A quantitative assessment of the effects of grazing on vegetation and soils over a global range of environments. Ecological Monographs 63:327–366. Milchunas, D. G., and W. K. Lauenroth. 1995. Inertia in plant community structure: State changes after cessation of nutrient-enrichment stress. Ecological Applications 5(2):452–458. Milchunas, D. G., and W. K. Lauenroth. 2001. Belowground primary production by carbon isotope decay and long-term root biomass dynamics. Ecosystems 4:139–150. Milchunas, D. G., W. K. Lauenroth, and I. C. Burke. 1998. Livestock grazing: Animal and plant biodiversity of shortgrass steppe and the relationship to ecosystem function. Oikos 83:65–74. Milchunas, D. G., W. K. Lauenroth, and P. L. Chapman. 1992. Plant competition, abiotic, and long- and short-term effects of large herbivores on demography of opportunistic species in a semiarid grassland. Oecologia 92:520–531. Milchunas, D. G., W. K. Lauenroth, P. L. Chapman, and M. K. Kazempour. 1989. Effects of grazing, topography, and precipitation on the structure of a semiarid grassland. Vegetatio 80:11–23. Milchunas, D. G., W. K. Lauenroth, P. L. Chapman, and M. K. Kazempour. 1990. Community attributes along a perturbation gradient in a shortgrass steppe. Journal of Vegetation Science 1:375–384. Milchunas, D. G., W. K. Lauenroth, and O. E. Sala. 1988. A generalized model of the effects of grazing by large herbivores on grassland community structure. The American Naturalist 132:87–106. Milchunas, D. G., A. R. Mosier, J. A. Morgan, D. LeCain, J. Y. King, and J. A. Nelson. 2005. CO2 and grazing effects on a shortgrass steppe: Forage quality versus quantity for ruminants. Agriculture Ecosystems and Environment 111:166–184. Milchunas, D. G., A. S. Varnamkhasti, W. K. Lauenroth, and H. Goetz. 1995. Forage quality in relation to long-term grazing history, current-year defoliation, and water resource. Oecologia 101:366–374. Morgan, J. A., D. R. LeCain, A. R. Mosier, and D. G. Milchunas. 2001. Elevated CO2 enhances water relations and productivity and affects gas exchange in C3 and C4 grasses of the Colorado shortgrass steppe. Global Change Biology 7:451–466. Mott, J. J. 1987. Patch grazing and degradation in native pastures of the tropical savannas in northern Australia, pp. 153–161. In: F. P. Horn, J. Hodgson, J. J. Mott, and R. W. Brougham (eds.), Grazing-lands research at the plant–animal interface. Winrock International, Morrilton, Alaska.

444

Ecology of the Shortgrass Steppe

Noy-Meir, I. 1988. Dominant grasses replaced by ruderal forbs in a vole year in undergrazed Mediterranean grasslands in Israel. Journal of Biogeography 15:579–587. Paine, R. T. 1966. Food web complexity and species diversity. The American Naturalist 100:65-75. Paine, R. T. 1971. A short-term experimental investigation of resource partitioning in a New Zealand rocky intertidal habitat. Ecology 52:1096–1106. Parton, W. J., M. P. Gutmann, and W. R. Travis. 2003. Sustainability and historical landuse change in the Great Plains: The case of eastern Colorado. Great Plains Research 3:97–125. Peterson, R. A. 1962. Factors affecting resistance to heavy grazing in needle-and-thread grass. Journal of Range Management 15:183–189. Pieper, R., C. W. Cook, and L. E. Harris. 1959. The effect of intensity of grazing upon nutritive content of diet. Journal of Animal Science 18:1031–1037. Rebollo, S., D. G. Milchunas, and I. Noy-Meir. 2005. Refuge effects of a cactus in grazed shortgrass steppe under different productivity, grazing intensity and cactus clump structure. Journal of Vegetation Science 16:85–92. Rebollo, S., D. G. Milchunas, I. Noy-Meir, and P. L. Chapman. 2002. The role of a spiny plant refuge in structuring grazed shortgrass steppe plant communities. Oikos 98:53–64. Reed, J. L., and D. D. Dwyer. 1971. Blue grama response to nitrogen and clipping under two soil moisture levels. Journal of Range Management 24:47–51. Ross, H. H. 1970. The ecological history of the Great Plains: Evidence from grassland insects, pp. 225–240. In: W. Dort and J. K. Jones (eds.), Pleistocene and recent environments of the central Great Plains. University Kansas Department Geology special publication no. 3. University of Kansas Press, Lawrence, Kans. Schlesinger, W. H., J. F. Reynolds, G. L. Cunningham, L. F. Huenneke, W. M. Jarrell, R. A. Virginia, and W. G. Whitford. 1990. Biological feedbacks in global desertification. Science 247:1043–1048. Schowalter, T. D., W. W. Hargrove, and D. A. Crossley, Jr. 1986. Herbivory in forested ecosystems. Annual Review of Entomology 31:177–196. Schwartz, C. C., and J. E. Ellis. 1981. Feeding ecology and niche separation in some native and domestic ungulates on the shortgrass prairie. Journal of Applied Ecology 18:343–353. Schwartz, C. C., J. G. Nagy, and R. W. Rice. 1977. Pronghorn dietary quality relative to forage availability and other ruminants in Colorado. Journal of Wildlife Management 41:161–168. Senft, R. L. 1989. Hierarchical foraging models: Effects of stocking and landscape composition on simulated resource use by cattle. Ecological Modeling 46:283–303. Senft, R. L., M. B. Coughenour, D. W. Bailey, L. R. Rittenhouse, O. E. Sala, and D. M. Swift. 1987. Large herbivore foraging and ecological hierarchies. BioScience 37:789–799. Senft, R. L., L. R. Rittenhouse, and M. A. Stillwell. 1984a. Diets selected by cattle from plant communities on shortgrass range. Proceedings of the Western Section of the American Society of Animal Science 35:180–183. Senft, R. L., L. R. Rittenhouse, and M. A. Stillwell. 1984b. Seasonal changes in nitrogen and energy budgets of range cattle. Proceedings of the Western Section of the American Society of Animal Science 35:200–203.

Effects of Grazing on Vegetation 445 Senft, R. L., L. R. Rittenhouse, and R. G. Woodmansee. 1985. Factors influencing patterns of cattle grazing behavior on shortgrass steppe. Journal of Range Management 38:82–87. Shaw, J. H. 1995. How many bison originally populated western rangelands? Rangelands 17:148–150. Shoop, M. C., R. C. Clark, W. A. Laycock, and R. M. Hansen. 1985. Cattle diets on shortgrass ranges with different amounts of fourwing saltbush. Journal of Range Management 38:443–449. Simoes, M., and Z. Baruch. 1991. Responses to simulated herbivory and water stress in two tropical C4 grasses. Oecologia 88:173–180. Sims, P. L., J. S. Singh, and W. K. Lauenroth. 1978. The structure and function of ten western North American grasslands. I. Abiotic and vegetational characteristics. Journal of Ecology 66:251–285. Sims, P. L., and J. S. Singh. 1978. The structure and function of ten western North American grasslands. II. Intra-seasonal dynamics in primary producer components. Journal of Ecology 66:547–572. Sims, P. L., D. W. Uresk, D. L. Bartos, and W. K. Lauenroth. 1971. Herbage dynamics on the Pawnee site: Aboveground and belowground herbage dynamics on the four grazing intensity treatments; and preliminary sampling on the ecosystem stress site. U.S. International Biological Program, Grassland Biome technical report no. 99. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. Singh, J. S., W. K. Lauenroth, H. W. Hunt, and D. M. Swift. 1984. Bias and random errors in estimation of net root production: A simulation approach. Ecology 65:1760–1764. Singh, J. S., D. G. Milchunas, and W. K. Lauenroth. 1998. Soil water dynamics and vegetation patterns in a semiarid grassland. Plant Ecology 134:77–89. Stebbins, G. L. 1981. Coevolution of grasses and herbivores. Annals of the Missouri Botanical Gardens 68:75–86. Uresk, D. W., and P. L. Sims. 1975. Influence of grazing on crude protein content of blue grama. Journal of Range Management 28:370–371. Van Vuren, D. R. 1987. Bison west of the Rocky Mountains: An alternative explanation. Northwest Science 61:65–69. Varnamkhasti, A. S., D. G. Milchunas, W. K. Lauenroth, and H. Goetz. 1995. Production and rain use efficiency in short-grass steppe: Grazing history, defoliation, and water resource. Journal of Vegetation Science 6:787–796. Vavra, M., R. W. Rice, R. M. Hansen, and P. J. Sims. 1977. Food habits of cattle on the shortgrass range in northeastern Colorado. Journal of Range Management 30:251–263. Vesey-Fritzgerald, D. 1972. Fire and animal impact on vegetation in Tanzania national parks. Tall Timbers Fire Ecology Conference 11:297–317. Vinton, M. A., D. C. Hartnett, E. J. Finck, and J. M. Briggs. 1993. Interactive effects of fire, bison (Bison bison) grazing and plant community composition in tallgrass prairie. American Midland Naturalist 129:10–18. Vokhiwa, Z. M. 1994. Carbon and nitrogen dynamics in grazed and protected semiarid shortgrass steppe. PhD diss., Colorado State University, Fort Collins, Colo. Weaver, J. W., and F. W. Albertson. 1956. Grasslands of the Great Plains: Their nature and use. Johnsen, Lincoln, Neb. Weaver, J. W., and F. E. Clements. 1938. Plant ecology. McGraw-Hill, New York.

446

Ecology of the Shortgrass Steppe

Westoby, M. 1985. Does heavy grazing usually improve the food resource for grazers? The American Naturalist 126:870–871. Westoby, M. 1986. Mechanisms influencing grazing success for livestock and wild herbivores. The American Naturalist 128:940–941. Whittaker, R. H. 1952. A study of foliage insect communities in the Great Smoky Mountains. Ecological Monographs 22:1–44. Williamson, S. C., J. K. Detling, J. L. Dodd, and M. I. Dyer. 1989. Experimental evaluation of the grazing optimization hypothesis. Journal of Range Management 42:149–152. Wilson, M. 1978. Archaeological kill site populations and the Holocene evolution of the genus Bison. Plains Anthropology 23:9–22.

17 Cattle Grazing on the Shortgrass Steppe Richard H. Hart Justin D. Derner

C

attle are the primary grazers on the shortgrass steppe. For example, during the late 1990s, 21 shortgrass counties in Colorado reported about 2.36 million cattle compared with 283,000 sheep (National Agricultural Statistics Service, USDA, 1997a), 60,000 pronghorn antelope, and a few thousand bison (Hart, 1994). Assuming one bison or five to six sheep or pronghorn consume as much forage as one bovine (Heady and Child, 1994), cattle provide about 97% of the large-herbivore grazing pressure in this region. The ratio of cattle to other grazers is even greater in the remainder of the shortgrass steppe. In 1997, the three panhandle counties of Oklahoma reported 387,000 cattle and only 1300 sheep, whereas the 38 panhandle counties of Texas reported 4.24 million cattle and 14,000 sheep (National Agricultural Statistics Service, USDA, 1997b,c). However, only about half the cattle in the panhandle counties of Texas and Oklahoma graze on rangeland the remainer are in feedlots.

Grazing Research on the Shortgrass Steppe Rangeland research on the shortgrass steppe (Table 17.1 describes the parameters of the major research stations in the shortgrass steppe) has included a long history of both basic ecology and grazing management. The responses of rangeland plant communities to herbivory are addressed by Milchunas et al. (chapter 16, this volume) and to disturbance are discussed by Peters et al. (chapter 6, this volume). Here we focus on research pertaining to three management practices important to cattle ranching on shortgrass steppe: stocking rates, grazing systems,

447

448

Ecology of the Shortgrass Steppe

Table 17.1 Parameters of Major Research Stations in the Shortgrass Steppe Location Central Plains Experimental Range, Nunn, Colorado Southeast Colorado Research Center, Springfield, Colorado Pantex IBP Site, Amarillo, Texas Texas Experimental Ranch, Lubbock, Texas

Latitude, Longitude

Elevation, m

Precipitation, mm

Frost-Free Days

40°50'N, 104°40'W 37°20'N, 102°40'W

1600–1700

325

135

1400

400

180

35°10'N, 101°50'W 33°40'N, 101°50'W

1170

495

200

1000

470

210

and extending the grazing season via complementary pastures and use of pastures dominated by Atriplex canescens [Pursh] Nutt (fourwing saltbush). Stocking Rates Stocking rate, defined as the number of animals per unit area for a specified time period, is the primary and most easily controlled variable in the management of cattle grazing. Cattle weight gain responses to stocking rate or grazing pressure (animal days per unit of forage produced) have been quantified in several grazing studies on the shortgrass steppe (Bement, 1969, 1974; Hart and Ashby, 1998; Klipple and Costello, 1960). Average daily gains per animal are better estimated as a function of grazing pressure, rather than stocking rate, as forage production is highly variable in this semiarid environment (Lauenroth and Sala, 1992; Milchunas et al., 1994). Average daily gain remains constant over a range of very low grazing pressures until a critical grazing pressure is reached; at this point, average daily gain declines linearly with increasing grazing pressure (Hart, 1972; Jones and Sandland, 1974). Decreases in average daily gains are the result of reduced nutrient intake as grazing pressure increases (Olson et al., 2002). The relation of gain to grazing pressure can be used to calculate profitability under a range of stocking rates, forage production levels, and cattle prices. In contrast to average daily gains, gains per unit area exhibit a quadratic function with stocking rate (Fig. 17.1). Economic returns per unit area depend not only on the relationships between gains and stocking rate, but also on selling price, maintenance, and operating costs. Maximum economic returns per unit area occur at stocking rates lower than those needed to maximize gain per unit land area (e.g., Hart et al., 1988; Manley et al., 1997). Relationships of cattle gains to stocking rate and grazing pressure for the shortgrass steppe have been developed from long-term experimental studies. The longest of these studies began in 1939 at the USDA–ARS Central Plains Experimental Range (CPER). This study encompasses light, moderate, and heavy

Relative gains or returns

Cattle Grazing on the Shortgrass Steppe 449 Daily gain

1

Gain/ha

0.8

Return/ha

0.6 0.4 0.2 0 0

0.2

0.4

0.6

0.8

1

Relative stocking rate or grazing pressure Figure 17.1

The classic stocking rate guide (modified from Bement, 1969).

grazing intensities, with four replications of 130-ha pastures at the beginning of the investigation. Replicates were removed from the study between 1950 and 1978, with only a single replicate pasture for each treatment remaining after 1978. From 1940 to 1964, light, moderate, and heavy grazing pastures were stocked with Hereford yearling heifers to remove 20%, 40%, and 60%, respectively, of the current year’s growth of grasses during a 5-month grazing season (mid May–mid October). From 1965 to the present, light, moderate, and heavy grazing pastures have been stocked to leave forage residual levels of 500, 335, and 225 kg⋅ha–1, respectively, at the end of the grazing season. The average stocking rates for the light, moderate, and heavy grazing treatments have been 15, 20, and 30 Hereford yearling heifers, respectively, in the 130-ha pastures over the 5-month grazing season. Of note, grazing pressure at each of the treatment intensity levels has nearly doubled from 1939 to 2006 (Derner et al., unpublished data), with this increase primarily a result of greater beginning grazing season weights of the Hereford yearling heifers, which are attributable to changes in calving seasons (from mid May–February and March), and genetic advances (e.g., artificial insemination). Grazing season gains from 1940 to 1949 were 129, 123, and 100 kg⋅head–1 under light, moderate, and heavy stocking, respectively (Klipple and Costello, 1960). Gains per unit area for this same time period were 13.0, 18.9, and 25.7 kg⋅ha–1 under light, moderate, and heavy stocking, respectively (Klipple and Costello, 1960). Primarily using these data, the seminal stocking rate guide for the shortgrass steppe was developed by Bement in 1969 (Fig. 17.1). Predicted maximum economic returns per unit area, calculated using cattle prices for 1964, 1965, and 1966, occurred when 336 kg⋅ha–1 of ungrazed forage was left at the end of the grazing season (Bement, 1974). Forage production, estimated as peak standing crop, was only determined in 17 of the years from 1940 through 1990. From 1991 to present, forage production has been determined annually. Using available data on forage production

450

Ecology of the Shortgrass Steppe

Mean daily weight gain (kg)

1.4 Heavy Moderate Light (calculated)

1.2 1.0 0.8 0.6 0.4 0.2 0

0 50 100 150 200 Grazing pressure (heifer-days Mg-1 peak standing crop)

Figure 17.2 Daily gain of yearling heifers under three grazing intensities and a range of grazing pressures at the USDA–ARS CPER, near Nunn, Colorado. (From Hart and Ashby [1998].)

and cattle gains, Hart and Ashby (1998) determined that average daily gains decrease linearly with increasing grazing pressure (Fig. 17.2), a relationship that can be used to predict cattle weight gain using a spreadsheet approach (Hart, 2000). Length of grazing season has a major effect on the response of cattle weight gains to grazing pressure (Hart and Hanson, 1993), because gains decrease rapidly as vegetative growth slows and stops, and nutrient concentrations decrease. Hyder et al. (1975) reported that gains declined sharply after the summer (June, July, August) maximum, even though forage availability remained high. Klipple and Costello (1960) found that heifers gained no weight in October under moderate stocking, and lost weight under heavy stocking rates. Yearling steer average daily gains decreased as stocking rate increased in an investigation conducted at the Southeastern Colorado Research Center (SEREC). Average daily gains were 0.75 kg ⋅ head–1 ⋅ day–1 under moderate and 0.69 kg ⋅ head–1 ⋅ day–1 under heavy stocking rates, at forage allowances of 285 and 215 kg ⋅ head–1 ⋅ 28 days–1, respectively (Cook and Rittenhouse, 1988). In 1974, the first year of the study, aboveground biomass production was 1280 kg⋅ha–1 under moderate and 1390 kg⋅ha–1 under heavy stocking rates. In 1980, the last year of the study, biomass production was 1240 and 1130 kg⋅ha–1 under moderate and heavy stocking rates, respectively. Grazing season was 168 days, from early May to late October. Grazing Systems Grazing systems have been developed as an alternative to season-long or continuous grazing. They typically involve dividing a larger pasture into smaller paddocks,

Cattle Grazing on the Shortgrass Steppe 451

and these small paddocks are grazed according to a schedule of use and grazing plan. Heady and Child (1994) described a grazing plan as one that stipulates the order and time at which pastures are to be grazed and rested. They also list several possible objectives of a grazing plan: 1. Improve range condition to maintain and increase plant vigor, to promote seed production and seedling establishment, to ensure vegetational succession, and/or to provide fuel for prescribed fire. 2. Improve quality and quantity of forage, and provide reserve forage for emergencies. 3. Achieve improved, regular distribution of grazing animals. 4. Promote uniform forage use by reducing selectivity of grazing. 5. Increase livestock weight gains and reproductive success, per head and/or per unit area, to increase ranch income. 6. Increase flexibility and decrease biological and financial risk in ranch operations. 7. Coordinate domestic animal grazing with habitat needs of wildlife and other uses of the land. Effects of time-controlled, short-duration rotational and season-long continuous grazing on yearling steer gains were evaluated from 1995 to 2003 at a moderate stocking rate (1.95 ha ⋅ animal unit–1⋅month–1) at the CPER. Steer average daily gains, grazing season gains, and beef production did not differ between grazing systems (Derner and Hart, 2007). Relationships between precipitation (annual or growing season) and average daily gain were not observed. In contrast, both grazing season gains (Fig. 17.3) and beef production exhibited a significant curvilinear response to both length of growing season and annual precipitation. Regression equations demonstrated that beef production is optimized when annual precipitation is 491 mm (24.8 kg⋅ha–1) and growing season precipitation (May–September) is 368 mm (25.1 kg⋅ha–1) (Derner and Hart, 2007). Average daily gains of yearling steers has been shown to be higher under continuous grazing (0.75 kg ⋅ head–1 ⋅ day–1) compared with a three-pasture rest–rotation grazing system (0.68 kg ⋅ head–1 ⋅ day–1) at SEREC from 1969 through 1977 (Cook and Rittenhouse, 1988). In this study, pastures were stocked to allow approximately 285 kg forage ⋅ steer –1 ⋅ 28 days–1. Grazing began about May 1 and ended about October 21, for a grazing season of 168 days. Mean aboveground biomass production was 1530 kg⋅ha–1 under continuous grazing and 1060 kg⋅ha–1 under rotation grazing. Eight-paddock rotation grazing and continuous season-long grazing were compared from 1983 to 1986 at SEREC using a stocking rate of 1 steer ⋅ 3.24 ha–1. Pastures were grazed from mid May until mid October, and aboveground biomass production was not estimated. Average daily gains of yearling steers was higher under continuous grazing (1.23 kg ⋅ head–1 ⋅ day–1) compared with the eight-paddock rotation grazing system (1.07 kg ⋅ head–1 ⋅ day–1) (Cook and Rittenhouse, 1988). Average daily gains for yearling steers were similar for continuous yearlong grazing and a rotation grazing system when stocking rates were the same at the Texas Experimental Ranch (Pitts and Bryant, 1987). When stocking rate was increased on the 16-paddock rotation grazing system, average daily gains

452

Ecology of the Shortgrass Steppe

Growing Season (May-September)

Grazing Season Gain (kg hd-1)

180 170 160 150 140 130

y = -137.63 + 1.774(x) - 0.0025(x2) p = 0.0497 r2 = 0.63

120 110 100

Continuous grazing Rotational grazing regression line

90 80 70

Grazing Season Gain (kg hd-1)

180

Annual

170 160 150 140 130 120

y = -181.94 + 1.474(x) - 0.0015(x2) p = 0.0426 r 2 = 0.65

110 100

Continuous grazing Rotational grazing regression line

90 80 70 150

200

250

300

350

400

450

500

550

600

Precipitation (mm) Figure 17.3 Relationships between grazing season gain and growing season precipitation (upper panel) and annual precipitation (lower panel) from 1995 to 2003 for season-long continuous and short-duration rotational grazing systems in the shortgrass steppe at the USDA–ARS CPER, near Nunn, Colorado. (From Derner and Hart [2007].)

decreased compared with continuous year-long grazing, but gains per unit land area increased (Table 17.2); this is the usual response to increased stocking rate or grazing pressure, as modeled by Hart (1978). Extending the Grazing Season Using Complementary Forages Supplemental hay, energy, and protein feedstuffs are fed to livestock in the shortgrass steppe from November until the next summer grazing season, because the

Cattle Grazing on the Shortgrass Steppe 453 Table 17.2 Stocking Rates, Average Daily Gain, and Gain per Hectare under 16-Paddock Short-Duration Rotation and Continuous Grazing on Shortgrass Rangeland, Texas Experimental Ranch, Lubbock, Texas Steers⋅ha⫺1

Gain, kg⋅ha⫺1

ADG, kg

Dates

SDG

CG

SDG

CG

SDG

CG

May 1979–Apr 1980 Apr 1980–Mar 1981 May 1981–Apr 1982 Apr 1982–Nov 1982

0.125 0.249 0.187 0.187

0.125 0.125 0.125 0.125

0.33 0.15 0.33 0.55

0.33 0.25 0.37 0.61

15.0 13.7 22.6 22.5

15.0 11.4 16.9 16.6

ADG, average daily gain; CG, continuous grazing; SDG, short-duration rotation grazing. Pitts and Bryant (1987).

wheat-fallow cropping regime predominantly used in this semiarid environment does not provide opportunities such as those in the more mesic environments in the Great Plains region, where grazing of crop residues can provide complementary forage during this time. Therefore, the efficiency of livestock production in the shortgrass steppe may be increased by grazing pastures that are seeded to cool-season perennial grasses and/or rangeland dominated by Atriplex canescens ([Pursh] Nutt) (fourwing saltbush) in late fall and/or early spring. Such pastures can complement native shortgrass rangeland by supplying forage in greater abundance and with greater quality during late fall and early spring, concurrent with periods of lower nutrient quality of warm-season grasses. Grazing of the saltbush-dominated rangeland may be deferred until after the summer months to extend the grazing season and to provide opportunities for additional animal gain via grazing. Atriplex canescens is adapted to semiarid conditions and produces palatable, digestible, high-protein forage (e.g., Cordova and Wallace, 1985; Garza and Fulbright, 1988; Rumbaugh et al., 1982). These shrubs typically begin rapid growth in May, flower during June, and complete seed set by the end of August. Trlica et al. (1977) determined that the most detrimental period for defoliating A. canescens is near maturity (i.e., flower production) compared with defoliation at quiescence, early growth, or rapid growth. Furthermore, Buwai and Trlica (1977) demonstrated that moderate defoliation (i.e., 60% removal of current year’s growth) during periods of rapid growth, seed set, or quiescence stimulated twig growth, but heavy defoliation (i.e., 90% removal) could kill plants. These authors did suggest, however, that use of A. canescens during either early spring or late fall may not adversely affect plants. Previous findings suggest that periodic rest is needed to maintain stable A. canescens populations (Pieper and Donart, 1978), as studies have indicated that continuous grazing results in marginal plant regrowth (Price et al., 1989), in reduced population densities (Schuman et al., 1990), and in increased proportion of nonflowering plants (Cibils et al., 2003). Shoop et al. (1985) estimated that A. canescens provided 32% of cattle diets on a pasture when its frequency was 19%, and 14% of diets on a pasture when its frequency was 8%. Although A. canescens plants that are continuously grazed by cattle produce little regrowth, plants entirely protected from grazing may also appear sometimes to

454

Ecology of the Shortgrass Steppe

produce little regrowth, because regrowth progressively decreases with increasing years of protection (Price et al., 1989). Weight gains of yearling heifers on A. canescens-dominated rangeland at the CPER are influenced by stocking rate for both the late fall (November–mid January) and early spring (April–mid May) grazing periods, but beef production is similar for both light and moderate stocking rates during each of the grazing periods (Derner and Hart, 2005). Similar beef production per unit land area between light and moderate stocking rates, both for late fall and early spring grazing periods, demonstrates that increased stocking rate does not compensate for lower individual animal gain that is achieved with the moderate stocking rate. Therefore, there is not an economic advantage to increase stocking rates from light to moderate levels when grazing A. canescens in the shortgrass steppe. These results, showing (1) adequate livestock gains with light stocking rates during both the late fall and early spring grazing periods and (2) that A. canescens plants clipped during either fall or spring exhibit equal recovery after the clipping just a few months later (Rumbaugh et al., 1982), suggest that incorporating light stocking rates not only provides reasonable beef production, but also is a sustainable land management practice in this ecosystem that does not degrade the rangeland resource. We have shown individual animal weight gains to be greater with light compared with moderate stocking rates for both of the grazing periods (Derner and Hart, 2005). Our findings suggest that land managers in the shortgrass steppe can effectively extend their grazing season by utilizing A. canescens-dominated rangeland both prior to and after the traditional summer grazing season using light stocking rates, which enhance individual animal gains without sacrificing gains per unit land area. Combining summer grazing of native rangeland with grazing of A. canescensdominated pastures and pastures seed to Psathrostachys juncea cv. Bozoisky (Bozoisky wildrye) and Agropyron cristatum × desertorum cv. Hycrest (Hycrest wheatgrass) results in a system in which cattle can graze for most of the calendar year, with the exception of the calving season, when cattle are commonly concentrated in small pastures near the ranch headquarters and fed hay or other stored feed. Research at the CPER demonstrates that weight gains of yearling heifers were 120 kg ⋅ head–1 on native shortgrass steppe during a 171-day summer grazing season, but 228 kg ⋅ head–1 when native rangeland was combined with seeded pastures and A. canescens-dominated pastures during a 312-day grazing period (start of April to end of January) (Hart, unpublished data).

Discussion and Conclusions Research and more than a century of experience have demonstrated that cattle grazing on the shortgrass steppe at light to moderate stocking rates is biologically sustainable, just as wild herbivore grazing was sustainable for millennia (Larson, 1940; Milchunas et al., chapter 16, this volume). Proper stocking rates are essential for economic sustainability; complementary forages must be evaluated for their contribution to economic sustainability. Grazing systems seem to confer few benefits

Cattle Grazing on the Shortgrass Steppe 455

to grazing cattle if other good management practices are followed. Climatic, atmospheric, economic, and sociopolitical changes may require compensatory changes in grazing management to ensure continued profitability and sustainability. Profitability of rangeland cattle grazing is threatened by economic, political, and social pressures outside the control of the cattle producer. Spitler (2001a) lists some of these factors: (1) the cyclical nature of the U.S. cattle industry; (2) high fixed costs and extreme price volatility; (3) until recently, decreasing per-capita demand for beef; (4) the beef industry’s failure to develop new easy-to-prepare products; and (5) exacerbation of oversupply by small producers not concerned with profits. In addition, four meat packers slaughter 85% to 90% of the cattle in the United States and thus exert great control over market prices (Spitler, 2001b). Spitler (2001a) urges ranchers to find ways to diversify their operation and to develop income sources unaffected by fluctuations of the conventional cattle market. The noneconomic roles of rangeland cattle grazing in preserving open space and habitat for other species, and the aesthetic value of open space to humans are often recognized. Of the total animal species found in the United States, 84% of the mammals, 74% of the birds, and 58% of the amphibians are found in nonforested rangelands, including but not limited to the shortgrass steppe (Hart, 1994). Several key threatened or endangered species are present as well (see other chapters in this book). Lastly, many observers would agree that extensive areas of shortgrass steppe rangeland, punctuated by herds of cattle and pronghorn antelope, have more aesthetic appeal than small-acreage housing tracts associated with overgrazed horse pastures. Land managers and livestock producers on the shortgrass steppe must balance profitability, stability, and sustainability, especially in light of high variability in precipitation amounts and resulting forage production (Derner and Hart, 2007; Lauenroth and Sala, 1992; Milchunas et al., 1994). A modeling approach for evaluating rangeland and livestock systems allows managers to assess the impacts of alternative management prior to actual implementation, thereby reducing risks in decision making. Models such as SMART (Hart, 1989), SPUR (Wright and Skiles, 1987), and SPUR2 (Hanson et al., 1992) have been developed to assist land managers. A new decision support system for the whole ranch–the Great Plains Framework for Agricultural Resource Management (GPFARM)—can serve as a tactical (short-term) and strategic (long-term) planning tool for production, economic, and environmental impact analysis, and site-specific database generation, from which alternative agricultural management systems can be tested and compared (Shafer et al., 2000). The GPFARM model has displayed excellent agreement in tracking growth and senescence trends in both warm- and cool-season perennial grasses, thereby demonstrating that it has functional utility for simulating forage production in semiarid environments such as the shortgrass steppe (Andales et al., 2005). References Andales, A. A., J. D. Derner, P. N. S. Bartling, L. R. Ahuja, G. H. Dunn, R. H. Hart, and J. D. Hanson. 2005. Evaluation of GPFARM for simulation of forage production and cow-calf weights. Rangeland Ecology and Management 58:247–255.

456

Ecology of the Shortgrass Steppe

Bement, R. E. 1969. A stocking-rate guide for beef production on blue-grama range. Journal of Range Management 22:83–86. Bement, R. E. 1974. Strategies used in managing blue-grama range on the Central Great Plains, pp. 160–166. In: K. W. Kreitlow and R. H. Hart (eds.), Plant morphogenesis as the basis for scientific management of range resources. USDA miscellaneous publication 1271. USDA, Washington, D.C. Buwai, M., and M. J. Trlica. 1977. Multiple defoliation effects on herbage yield, vigor and total nonstructural carbohydrates of five range species. Journal of Range Management 30:164–171. Cibils, A. F., D. M. Swift, and R. H. Hart. 2003. Changes in shrub fecundity in fourwing saltbush browsed by cattle. Journal of Range Management 56:39–46. Cook, C. W., and L. R. Rittenhouse. 1988. Grazing and seeding research at the Southeastern Colorado Experiment Station, Springfield, CO. Colorado State University Agricultural Experiment Station technical bulletin LTB88–8. Colorado State University Agricultural Experiment Station , Fort Collins. Cordova, F. J., and J. Wallace. 1985. Nutritive value of some browse and forb species. Proceedings of American Society of Animal Science 26:160–162. Derner, J. D., and R. H. Hart. 2005. Heifer performance under two stocking rates on fourwing saltbush-dominated rangeland. Rangeland Ecology and Management 58:489–494. Derner, J. D., and R. H. Hart. 2007. Livestock and vegetation responses to rotational grazing in shortgrass steppe. Western North American Naturalist 67:359–367. Garza, A., Jr., and T. E. Fulbright. 1988. Comparative chemical composition of armed saltbush and fourwing saltbush. Journal of Range Management 41:401–403. Hanson, J. D., B. B. Baker, and R. M. Bourdon. 1992. SPUR2 documentation and user guide. GPSR technical report no. 1. USDA–ARS, Fort Collins, Colo. Hart, R. H. 1972. Forage yield, stocking rate, and beef gains on pasture. Herbage Abstracts 42:345–353. Hart, R. H. 1978. Stocking rate theory and its application to grazing on rangelands, pp. 550–553. In: D. N. Hyder (ed.), Proceedings of the First International Rangeland Congress. Society for Rangeland Management, Denver, Colo. Hart, R. H. 1989. SMART: A simple model to assess range technology. Journal of Range Management 42:421–424. Hart, R. H. 1994. Rangeland. In: Charles Arntzen (ed.), Encyclopedia of Agriculture Science 3491-–501. Academic Press, Inc., San Diego, CA. Hart, R. H. 2000. The right rate: New spreadsheets help producers estimate most profitable stocking rates. Western Farmer-Stockman 120(11):WB4, WB13, WB14. Hart, R. H., and M. M. Ashby. 1998. Grazing intensities, vegetation, and heifer gains: 55 years on shortgrass. Journal of Range Management 51:392–398. Hart, R. H., and J. D. Hanson. 1993. Managing for economic and ecological stability of range and range-improved grassland systems with the SPUR II model and the STEERISKIER spreadsheet. Proceedings of the XVII International Grassland Congress, Palmerston North. 8–21 February 1993:1593–1598. Hart, R. H., M. J. Samuel, P. S. Test, and M. A. Smith. 1988. Cattle, vegetation and economic responses to grazing systems and grazing pressure. Journal of Range Management 41:282–286. Heady, H. F., and R. D. Child. 1994. Rangeland ecology and management. Westview Press, Boulder, Colo. Hyder, D. N., R. E. Bement, E. E. Remmenga, and D. F. Hervey. 1975. Ecological responses of native plants and guidelines for management of shortgrass range. USDA technical bulletin 1503. U.S. Government Printing Office, Washington, D.C.

Cattle Grazing on the Shortgrass Steppe 457 Jones, R. J., and R. L. Sandland. 1974. The relation between animal gain and stocking rate. Derivation of the relations from the results of grazing trials. Journal of Agricultural Science 83:335–342. Klipple, G. E., and D. F. Costello. 1960. Vegetation and cattle responses to different intensities of grazing on shortgrass ranges on the Central Great Plains. USDA technical bulletin 1216. U.S. Government Printing Office, Washington, D.C. Larson, F. 1940. The role of bison in maintaining the shortgrass plains. Ecology 21:113–121. Lauenroth, W. K., and O. E. Sala. 1992. Long term forage production of North American shortgrass steppe. Ecological Applications 2:397–403. Manley, W. A., R. H. Hart, M. J. Samuel, M. A. Smith, J. W. Waggoner, Jr., and J. T. Manley. 1997. Vegetation, cattle and economic responses to grazing strategies and pressures. Journal of Range Management 50:638–646. Milchunas, D. G., J. R. Forwood, and W. K. Lauenroth. 1994. Productivity of long term grazing treatments in response to seasonal precipitation. Journal of Range Management 47:133–139. National Agricultural Statistics Service, USDA. 1997a. 1997 Census of agriculture. Vol. 1, part 6, Colorado. State & county data. U.S. Government Printing Office, Washington, D.C. National Agricultural Statistics Service, USDA. 1997b. 1997 Census of agriculture. Vol. 1, part 36, Oklahoma. State & county data. U.S. Government Printing Office, Washington, D.C. National Agricultural Statistics Service, USDA. 1997c. 1997 Census of agriculture. Vol. 1, part 43, Texas. State & county data. U.S. Government Printing Office, Washington, D.C. Olson, K. C., J. R. Jaeger, J. R. Brethour, and T. B. Avery. 2002. Steer nutritional response to intensive-early stocking on shortgrass rangeland. Journal of Range Management 55:222–228. Pieper, R. D., and G. B. Donart. 1978. Response of fourwing saltbush to periods of protection. Journal of Range Management 31:314–315. Pitts, J. S., and F. C. Bryant. 1987. Steer and vegetation response to short duration and continuous grazing. Journal of Range Management 40:386–389. Price, D. L., G. B. Donart, and G. M. Southward. 1989. Growth dynamics of fourwing saltbush as affected by different grazing management systems. Journal of Range Management 42:158–162. Rumbaugh, M. D., D. A. Jounson, and G. A. Van Epps. 1982. Forage yield and quality in a Great Basin shrub, grass, and legume pasture experiment. Journal of Range Management 35:604–609. Schuman, G. E., D. T. Booth, and J. W. Waggoner. 1990. Grazing reclaimed mined land seeded to native grasses in Wyoming. Journal of Soil and Water Conservation 45:653–657. Shafer, M. J., P. N. S. Bartling, and J. C. Ascough, II. 2000. Object-oriented simulation of whole farms: GPFARM framework. Computers and Electronics in Agriculture 28:29–49. Shoop, M. C., R. C. Clark, W. A. Laycock, and R. M. Hansen. 1985. Cattle diets on shortgrass ranges with different amounts of fourwing saltbush. Journal of Range Management 38:443–449. Spitler, J. 2001a. Spread your risk. Western Farmer-Stockman 121(2):WB4–7. Spitler, J. 2001b. Utah hard times. Western Farmer-Stockman 121(2):WB1–2.

458

Ecology of the Shortgrass Steppe

Trlica, M. J., M. Buwai, and J. Menke. 1977. Effects of rest following defoliations on the recovery of several range species. Journal of Range Management 30:21–27. Wright, J. R., and J. W. Skiles (eds.). 1987. SPUR: Simulation of production and utilization of rangelands. Documentation and user guide. USDA-ARS 63. National Technical Information Service, Springfield, Va.

18 Effects of Grazing on Abundance and Distribution of Shortgrass Steppe Consumers Daniel G. Milchunas William K. Lauenroth

A

lthough livestock are the most obvious consumers on the shortgrass steppe, they are certainly not the only consumers. However, livestock may influence the other consumers in a number of different ways. They may directly compete for food resources with other aboveground herbivores. There is behavioral interference between livestock and some species of wildlife (Roberts and Becker, 1982), but not others (Austin and Urness, 1986). The removal of biomass by livestock alters canopy structure (physiognomy) and influences microclimate. Bird, small-mammal, and insect species can be variously sensitive to these structural alterations (Brown, 1973; Cody, 1985; MacArthur, 1965; Morris, 1973; Rosenzweig et al., 1975; Wiens, 1969). There are both short- and long-term effects of grazing on plant community species composition, primary production, and plant tissue quality. Belowground consumers can also be affected by the effects of grazing on soil water infiltration, nutrient cycling, carbon allocation patterns of plants, litter accumulation, and soil temperature. The overall effects of livestock on a particular component of the native fauna can be negative or can be positive through facilitative relationships (Gordon, 1988). In this chapter we assess the effects of cattle grazing on other above- and belowground consumers, on the diversity and relative sensitivity of these groups of organisms, and on their trophic structure. We first present some brief background information on plant communities of the shortgrass steppe and on the long-term grazing treatments in which many of the studies reported herein were conducted. Details on the plant communities are presented by Lauenroth in chapter 5 (this volume), grazing effects on plant communities by Milchunas et al. in chapter 16 (this volume); and grazing effects on nutrient distributions and cycling by Burke et al. in chapter 13 (this volume). 459

460

Ecology of the Shortgrass Steppe

Plant Communities and Grazing Intensities The physiognomy of the shortgrass steppe is indicated in its name. The dominant grasses (Bouteloua gracilis and Buchloë dactyloides), forb (Sphaeralcea coccinea), and carex (Carex eleocharis) have the majority of their leaf biomass within 10 cm of the ground surface. A number of less abundant midheight grasses and dwarf shrubs are sparsely interspersed among the short vegetation, but usually much of their biomass is within 25 cm of the ground. Basal cover of vegetation typically totals 25% to 35%, and is greater in long-term grazed than in ungrazed grassland. Bare ground (more frequent on grazed sites) and litter-covered ground (more frequent on ungrazed sites) comprise the remainder of the soil surface (Milchunas et al., 1989). On sandy soils, the true shrub Atriplex canescens provides added structural complexity, although understory dominants remain similar (Fig. 18.1A). Consumer populations can differ substantially between the shrubland and grassland communities, probably because of differences in structure of canopy cover. Topography also influences consumer populations if they preferentially use either the uplands (catena ridgetops) or lowlands (catena toe slopes, also referred to as swales). Grazing by cattle is much greater in lowlands, and this tends to homogenize differences in plant community species composition that occur in the absence of grazing resulting from soil differences (Milchunas et al., chapter 16, this volume). Often, less extensive community types are associated with particular soils (Lauenroth, chapter 5, this volume). The shortgrass steppe is in a semiarid environment, with annual precipitation averaging only 321 mm · y–1 at the Central Plains Experimental Range (CPER). Aboveground net primary production (ANPP) averages 95 g · m–2 in moderately grazed uplands, with slightly greater amounts in lowlands compared with uplands and in shrublands compared with grasslands (Lauenroth et al., chapter 12, this volume). Long-term grazing treatments were established at the CPER in 1939. Light, moderate, and heavy grazing treatments result in approximately 20%, 40%, and 60% removal of ANPP when averaged over years (see Milchunas et al., chapter 16, this volume, for details of stocking rates and different annual removal rates). Pastures not in the more tightly regulated treatment pastures are generally grazed at moderate levels, whereas grasslands in the adjacent Pawnee National Grassland (PNG) or in private ownership are more often grazed at 50% to 65% removal of ANPP. The long-term grazing treatments at the CPER are grazed either summer or winter only. Other grazing systems may be used at other locations in the region.

Large Herbivores Pronghorn (Antilocapra americana, sometimes referred to as antelope) are the most abundant large native herbivore in the shortgrass steppe (Fig. 18.1B). Numbers in eastern Colorado were estimated to be 10,000 in 1958 (Hoover et al., 1959) and 60,000 in 1992 (Hart, 1994), compared with two million presettlement

(A)

(B)

Figure 18.1 (A) Pronghorn antelope on a level uplands site. (Photo from Natural Resource Ecology Lab archives.) (B) An Atriplex shrubland community ungrazed since 1939 left of fence line and moderately grazed by cattle right of fence line. (Photo by Mark Vandever.) Note shrub abundance is lower in the ungrazed treatment.

462

Ecology of the Shortgrass Steppe

and a low of 1000 in 1918 (Hoover et al., 1959). Lauenroth and Milchunas (1991) calculated a presettlement biomass of pronghorn to have been 16 kg · ha–1 compared with the current cattle biomass of 76 kg · ha–1 in heavily grazed pastures. Current numbers are regulated by hunting, and this sometimes depends upon problems associated with pronghorn grazing of winter wheat crops. Across their range, pronghorn numbers consistently increased from the early 1970s through the early 1990s (Langner and Flather, 1994). Although mule deer (Odocoileus hemionus) and white-tailed deer (O. virginianus) are not common on the open plains, they can often be found near ravines and riparian/shrubland areas where cover is available. There are no longer any free-ranging bison (Bison bison), but they are found in increasing numbers in some natural areas and parks where populations are being restored, and on bison ranches, which are also increasing in numbers. The long evolutionary history of grazing by these native ungulates in the shortgrass steppe is discussed by Hart in chapter 4 (this volume). Bison and cattle are closely related and similar in size. This raises questions concerning the effect of the replacement of bison with cattle on other native consumers and the compatibility of the two in areas where bison are being reintroduced. Grasses and sedges are a large proportion of the diet of both species. Relatively minor differences in their diets have been found in the shortgrass steppe (Peden et al., 1974; Schwartz and Ellis, 1981), in the Great Basin shrub steppe in Utah (Van Vuren, 1982), and the mixed-grass prairie of South Dakota (Plumb and Dodd, 1993). In the shortgrass steppe, cattle consume more cool-season grasses, forbs, and shrubs than bison (Schwartz and Ellis, 1981). Bison diets have been found to be largely warm-season grasses, and they are somewhat less selective feeders than cattle. The low degree of selectivity by bison is reflected in the poor quality of their diets, and in higher intake as a proportion of their body weight (75 g · kg–1wt0.75) relative to cattle (52 g · kg–1 wt0.75) or sheep (46 g · kg–1 wt0.75) (Rice et al., 1974). There is some spatial niche separation between cattle and bison, in that cattle spend more time during the growing season grazing in swales (Peden et al., 1974; Van Vuren, 1982). Bison and cattle are socially compatible down to distances of about 4 m, past which cattle are subordinate to bison (Van Vuren, 1982). Observation of a mixed herd in lightly and heavily grazed shortgrass steppe has shown that bison are dominant over cattle in selection of grazing and bedding areas and in obtaining water (Sparks, 1972). No differences are apparent in time spent grazing or resting, although both species spend more time grazing in the heavily than lightly grazed pasture. The high dietary overlap between cattle and bison, however, suggests a high degree of competition, but also a generally similar pressure on the plant community where cattle have replaced the native bison. There are some minor differences between cattle and bison grazing. Cattle are a little more selective than bison for components much more common in pronghorn diets (forbs and shrubs), and these components are more susceptible to fluctuating abiotic conditions. This led Lauenroth and Milchunas (1991) to conclude that a bison–pronghorn assemblage was probably more stable than a pronghorn–cattle one. However, Schwartz et al. (1977) found that pronghorn were able to maintain similar diet quality on long-term lightly and heavily cattle-grazed pastures. This

Effects of Grazing on Abundance and Distribution of Consumers 463

may be the result of a capacity for high selectivity in their foraging, differences in preference for topographic positions across the landscape (Schwartz and Ellis, 1981), and the tendency for cattle to be less mobile and restricted in their foraging areas (Van Vuren, 1982). However, competition between species may only occur during uncommon bottleneck periods, when all food becomes scarce. An example of this is a pronghorn die-off that occurred during a time of drought in Texas, and was attributed to overgrazing by domestic animals (Hailey et al., 1966). Dietary overlap between pronghorn and sheep is much greater than between pronghorn and cattle (Clary and Holmgren, 1982; Schwartz and Nagy, 1976). Although social avoidance of sheep by pronghorn has not been observed, pronghorn in the Great Basin avoid areas grazed by sheep during winter until spring regrowth occurs, and favor areas temporarily rested from sheep use (Clary and Beale, 1983; Clary and Holmgren, 1982).

Small Mammals Lagomorphs and rodents are an important component of the native mammalian fauna of the shortgrass steppe. However, Lauenroth and Milchunas (1991) estimated that consumption by lagomorphs and rodents combined was less than 1% of ANPP in the northern shortgrass steppe, even though each of their biomasses was greater than that of pronghorn antelope. Rodents process approximately 40,000 kJ · ha–1 · y–1 via herbivory compared with 96,000 kJ · ha–1 · y–1 as predators, whereas lagomorphs are 100% herbivores. The ratio of rodent to lagomorph biomass was much greater at the southern shortgrass site (Texas) compared with the northern site (Colorado). The ecology of lagomorphs and rodents in the shortgrass steppe is addressed by Stapp et al. in chapter 8 (this volume). The effects of livestock grazing on small mammals can manifest through direct dietary competition, by altering forage quality and seed or arthropod abundances, and by changing community physiognomy and predation rates. Lagomorphs Grasses and sedges represent more than 50% of the diet of black- and white-tailed jackrabbits (Lepus californicus, L. townsendi) and desert cottontails (Sylvilagus auduboni) (Flinders and Hansen, 1972; Hansen and Gold, 1977). Agropyron smithii, a grass species highly preferred by cattle, is also a very important component in the diet of all three lagomorphs. Both cattle and black-tailed jackrabbits have been found to graze swales preferentially during the growing season (Flinders and Hansen, 1975; Milchunas et al., chapter 16, this volume), but the white-tailed jackrabbit prefer uplands, and cottontails show no preference. There is potential for overlap between cattle and lagomorphs in both diet and habitat preference. Flinders and Hansen (1975) found that black-tailed jackrabbits are more than twice as abundant in lightly and moderately summer-grazed than in heavily summer-grazed or moderately winter-grazed treatments. They noted that this is

464

Ecology of the Shortgrass Steppe

not the usual response of black-tailed jackrabbits to grazing, because in more productive habitats they tend to avoid dense vegetation and prefer heavily grazed or burned communities. This has been observed in the more productive grasslands of eastern Texas (Taylor and Lay, 1944), the sand hills of Colorado (Sanderson, 1959), southern Arizona (Taylor et al., 1935), and the mixed-grass (Smith, 1940) and tallgrass prairie (Phillips, 1936) of Oklahoma. Rabbit and rodent numbers become more abundant as range condition deteriorates in desert grassland, leading to increased competition with livestock (Schmutz et al., 1992). Heavy grazing of short-stature vegetation may reduce cover or forage to less than a critical minimum. White-tailed jackrabbits show no significant differences between grazing treatments in the shortgrass steppe, but average densities are highest in moderately grazed treatments (Flinders and Hansen, 1975). Cottontails are significantly more abundant in moderately grazed treatments compared with those either lightly or heavily grazed. Cottontails appear to be closely associated with shrubs, and moderate grazing of the shrubs seems to increase adventitious budding and the development of a denser shrub canopy. Prairie Dogs Prairie dogs (Cynomys ludovicianus) have an important influence on both the flora and fauna of Great Plains grasslands. Potentially 170 vertebrate species are closely associated with prairie dog colonies (Miller et al., 1994), and plant diversity is increased in some communities as a result of their burrowing and clipping activities (Bohnam and Lerwick, 1976; Hansen and Gold, 1977; Whicker and Detling, 1988). Millions of prairie dogs occupied the Plains before settlement, and a single colony in Texas was reported to have covered 6.5 million ha (Merriam, 1902). Large-scale poisoning throughout their range, because of perceived competition with cattle and their potential as disease vectors, has drastically reduced numbers and sizes of towns (McNulty, 1971). Dietary overlap between cattle and prairie dogs is relatively high in the shortgrass steppe, ranging from a high of 69% in spring to a low of 41% in winter (Hansen and Gold, 1977). Hansen and Gold (1977) concluded that prairie dogs are not, however, influenced by cattle grazing in the shortgrass steppe. This is not the usual response to grazing. Prairie dogs prefer heavily grazed areas in more productive grasslands than the shortgrass steppe, and can decline in numbers if the cattle are removed (Allen and Osborn, 1954; Knowles, 1986; Koford, 1958). Prairie dogs may either positively or negatively impact cattle. In the shortgrass steppe, Hansen and Gold (1977) estimated that forage consumption and soil disturbances by prairie dogs and the associated cottontails reduced the amount of ANPP available to cattle by 24%. Cattle either gained no weight or lost weight during winter in pastures with prairie dog towns. O’Meilia et al. (1982) similarly reported that weight losses appeared to occur only during fall and winter in B. gracilis-dominated grassland in Oklahoma. The lack of weight gain differences during the growing season between pastures with and without prairie dog towns was attributed to potentially greater forage quality on the towns. Knowles (1986) observed that cattle occurred significantly more often, and Hassien (1976)

Effects of Grazing on Abundance and Distribution of Consumers 465

found greater numbers of fecal pats, in areas with prairie dog colonies as opposed to those without colonies. This suggests that prairie dogs facilitate grazing by cattle in a manner similar to that reported for bison (Whicker and Detling, 1988). Greater mineralization and lower immobilization rates in prairie dog towns compared with uncolonized grassland resulted in increased availability of soil nitrogen and increased shoot nitrogen concentrations in mixed-grass prairie (Holland and Detling, 1990). The defoliation activities of the prairie dogs also lowered stem-to-leaf ratios and increased digestibilities of plants for bison in the mixedgrass prairie (Coppock et al., 1983). Recent studies from the shortgrass steppe in Colorado show that the facilitation effect of prairie dogs on large herbivores does not necessarily occur as it does in more productive plant communities, and that weight gains of cattle are negatively affected by growing-season grazing. Observations of grazing behavior of cattle indicate no preference for prairie dog towns compared with adjacent offtown locations (Guenther and Detling, 2003). This suggests the quality of forage did not differ enough to alter grazing behavior. Grazing by prairie dogs in more productive communities is likely to have greater effects on leaf-to-stem ratios and on plant species composition that it does in short, less productive plant communities (Milchunas et al., 1988). Cattle weight gains declined 5.5% when the percentage of pastures occupied by prairie dogs was 20% compared with no towns, and was reduced 13.9% when the percentage occupied was 60% (Derner et al., 2006). The somewhat low weight gain losses compared with the percentage area occupied was attributed to the grazing tolerance of the dominant shortgrasses. This study was on areas recently experiencing large increases in prairie dog towns, and greater impacts may be expected in areas occupied for longer times. Small Rodents Rodent populations are often correlated with various structural attributes of plant communities, particularly cover (Brown, 1973; French et al., 1976; Grant and Birney, 1979; Rosenzweig et al., 1975; Stapp et al., chapter 8, this volume). Rodents may be important consumers of seeds and/or insects (Flake, 1971; French et al., 1976). Competitive interactions between rodent species influence the composition of rodent communities (Valone and Brown, 1995). Grazing can alter plant community structure, the abundance of insects, and seed production. These alterations may affect competition and/or social interactions between rodent species. In the shortgrass steppe, grazing results in large reductions in aboveground arthropod abundances (Crist, chapter 10, this volume), decreased seed production of B. gracilis (Coffin and Lauenroth, 1992), but increased seed production in some species (Rebollo et al., 2002). Possibly because of the relatively small changes in community structure and plant species composition with grazing in this system, there are only small effects of grazing on rodent communities in the shortgrass steppe compared with those found in more productive communities (Grant et al., 1982). For example, small-mammal production has been estimated to be 346 and 486 kcal · ha–1 in lightly and heavily grazed shortgrass steppe compared with 5812 and 1004 kcal · ha–1 in ungrazed and grazed tallgrass prairie. The

466

Ecology of the Shortgrass Steppe

primarily surface-dwelling, granivorous and omnivorous species of the shortgrass steppe are adapted to open habitat (Stapp et al., chapter 8, this volume). Although rodent productivity was found to be somewhat greater, rodent biomass was slightly lower in heavily (160 g live weight · ha–1) than in lightly (175 g live weight · ha–1) grazed shortgrass steppe over a 4-year study (Grant et al., 1982). Only deer mice (Peromyscus maniculatus) increased with heavy grazing. In a 2-year study of three grazing intensity treatments, Flake (1971) found no consistent relationship across treatments in abundances of Ord’s kangaroo rats (Dipodomys ordii), more northern grasshopper mice (Onychomys leucogaster) in moderately grazed than in either lightly or heavily grazed treatments, increases of deer mice with increasing grazing intensity, and decreases of thirteen-lined ground squirrels (Spermophilus tridecemlineatus) with increasing grazing intensity. Pocket gophers (Thomomys bottae) did not appear to select habitat based upon grazing treatments. Mounds averaged 6.5%, 2.5%, and 8% of land cover in the ungrazed, lightly, and heavily grazed treatments, respectively (Grant et al., 1980).

Birds Similar to rodents, birds often respond greatly to alterations in vegetation canopy structure (Cody, 1985; MacArthur, 1965; Wiens, 1969). However, vertical layering of habitat structure in grasslands is not as great a potential control on bird diversity as it would be in forests, and birds do not appear to respond to differences in horizontal habitat heterogeneity across or within 15 sites and five community types (Wiens, 1974a,b; Wiens and McIntyre, chapter 9, this volume). Furthermore, with respect to food resources, data from four grassland and shrub steppe communities suggests that food may not usually limit bird populations, although niche overlap may increase in less productive habitats (Wiens and McIntyre, chapter 9, this volume; Wiens and Rotenberry, 1979). Therefore, one may expect the response of birds to grazing in the shortgrass steppe to be similar to that reported for small rodents. However, this is not the case. A study of nesting birds in the long-term grazing treatments in the northern shortgrass steppe showed a complete shift in the dominant species from Lark Bunting (Calamospiza melanocorys) to Horned Lark (Eremophila alpestris), and very low similarity in species composition between lightly and heavily grazed treatments (Giezentanner, 1970). Utilization of heavily, moderately, and lightly summer-grazed, upland grassland communities averaged 3805, 2446, and 2789 bird-use days (40/ha) in summer, respectively, and 3680, 627, and 1945 bird-use days (40/ha) in winter. Estimates for winter-grazed, lowland shrubland communities were 2384, 2438, and 2114 bird-use days (40/ha) in summer; and 1171, 1263, and 541 bird-use days (40/ha) in winter for the heavily, moderately, and lightly grazed treatments, respectively. In a 5-year study, Ryder (1980) found that nesting by Mountain Plover (Eupoda montana), Horned Lark, and McCown’s Longspur (Rhynchophanes mccownii) was greater on heavily than lightly grazed pastures. In contrast, Chestnut-collared Longspur (Calcarius ornatus), Western

Effects of Grazing on Abundance and Distribution of Consumers 467

Meadowlark (Sturnella neglecta), and Lark Bunting nesting were all greatest on the lightly grazed treatment. A study at a southern shortgrass site in Texas, however, shows very little difference in bird species composition with grazing. Wiens (1973) calculated a 98% similarity between grazed and ungrazed bird communities, and 96% similarity between plant communities. Abundance of birds was, however, greater in heavily (7.2 kg · 100 ha–1) than in lightly (2.4 kg · 100 ha–1) grazed shortgrass steppe sites at Muleshoe National Wildlife Refuge in west Texas (Grzybowski, 1980, 1982). McCown’s Longspurs were found only in the heavily grazed treatment. Horned Larks and Chestnut-collared Longspurs showed large increases with grazing, and Baird’s Sparrow (Ammodramus bairdii) showed decreases. Similarity between the grazing treatments in species composition was 74%. The response of a particular bird species to grazing is not necessarily consistent across sites of different productivities. Depending on the site, grazing may lead toward or away from some vegetation density seen as optimum by a given bird species (Bock and Webb, 1984; Kantrud and Kologiski, 1982). Endemic grassland birds that are closely associated with the shortgrass steppe often prefer open, sparse vegetation (Fig. 18.2 [Knopf, 1996]). Although endemic grassland birds of North America are a threatened group, this is more likely the result of loss of grassland to cultivation and development, and of conditions on the wintering grounds than a result of grazing by domestic livestock. All endemic grassland birds evolved with grazing by large native ungulates, with the exception of Cassin’s Sparrow (Aimophila cassinii), whose southwest distribution is beyond the historical range of bison (Bock and Webb, 1984; Knopf, 1996). The Mountain Plover is a category I threatened species (major declines have been documented), and this species selects nesting sites that are in very sparse vegetation. Graul (1973, 1975) and Ryder (1980) found Mountain Plovers nesting almost exclusively on very heavily grazed shortgrass steppe. Plovers have been observed to nest on plowed ground (Shackford, 1991), in the middle of two-track roadways through

Grazing Intensity Excessive < -- Heavy ---------- Moderate --------- Light ------- > None Mountain Plover McCown’s Longspur Ferruginous Hawk Long-billed Curlew Lark Bunting Chestnut-collared Longspur Sprague’s Pipit Baird’s Sparrow

Cassin’s Sparrow

Bare Mixed/Shrub

Plant Community Structure Figure 18.2 Distributions of endemic grassland bird species in relation to grazing intensity and plant community structure. (From Knopf [1996].)

468

Ecology of the Shortgrass Steppe

pastures (Milchunas, personal observation), and alongside cattle fecal pats, which provide both protection from wind and a source of insect food. This illustrates the close association of Mountain Plovers with the large herds of bison that “ . . . so completely consumed the herbage of the plains that detachments of the United States Army found it difficult to find sufficient grass for their mules and horses” (Hornaday, 1889 [quoted by Larson, 1940, p. 117]). Nighthawks (Chordeiles minor), Killdeer (Charadrius vociferus) (Ryder 1980), and McCown’s Longspurs (Knopf, 1996) also prefer heavily grazed habitat throughout the PNG. Long-billed Curlews (Numenius americanus; category II threatened candidate) prefer grazed areas in Comanche National Grasslands in southern Colorado (King, 1978) and short vegetation throughout its range in western North America (Paton and Dalton, 1994). However, nest losses to predators were greater in heavily than in moderately grazed pastures (With, 1994). The response of raptors to different grazing intensities is difficult to assess because of their wide-ranging habits. The influence of grazing is probably related to some extent to the impact on their primary food source of lagomorphs and rodents. Indirect effects can include cover for prey, nesting sites, and human management activities. The Ferruginous Hawk (Buteo regalis; category II threatened candidate) prefers hunting in grazed areas because the reduced cover affords better detection of prey (Wakeley, 1978). The Ferruginous Hawk is the only raptor to nest sometimes on the ground, and all raptors are nest-site limited in this primarily open grassland habitat (Olendorff and Stoddart, 1974). Nesting sites created by humans may have positive impacts on raptor populations, and percentages of these types of nesting situation were 0%, 4%, 34%, and 52% for Prairie Falcon (Falco peregrinus), Golden Eagle (Aquila chrysatos), Swainson’s Hawk (B. swainsoni), and Ferruginous Hawk, respectively (Olendorff, 1972). Human control of prairie dog populations can negatively affect raptors by reducing a food source important to them. Raptor numbers have been closely tied to the presence or absence of active prairie dog colonies (Cully, 1991).

Arthropods Approximately 90% of arthropod consumption occurs belowground in the shortgrass steppe (Crist, chapter 10, this volume; Lauenroth and Milchunas, 1991). Arthropods are an important herbivore group, accounting for approximately one third of total and 14% of aboveground herbivory. Herbivores comprise approximately 85% of all macroarthropods and 28% of all microarthropods. Macroarthropods Grazing has very large impacts on both aboveground and belowground macroarthropods in terms of both biomass and numbers. Aboveground macroarthropods averaged 42, 51, 32, and 32 · m–2, and belowground macroarthropods averaged 183, 151, 87, and 119 · m–2, in the ungrazed, lightly, heavily grazed, and currentyear ungrazed (previously heavily grazed) treatments, respectively (Andrews,

Effects of Grazing on Abundance and Distribution of Consumers 469

1977). The lack of an aboveground response to current-year removal of grazers from the heavily grazed treatment suggests that the longer term effects of the grazers may be more important than the immediate loss of plant biomass to consumption. Productivity of macroarthropods in relation to grazing treatments follows a pattern similar to that for density, and all trophic groups were somewhat similarly affected by ungrazed, lightly, and heavily grazed treatments (Fig. 18.3 [Milchunas et al., 1998]). The decline in the proportion of belowground omnivores and saprophages with increasing grazing intensity is probably not significant, because values for these relatively small groups belowground were quite variable. The lack of a proportionately greater decline in herbivores is surprising, given the potential for direct competition with cattle. Slight increases in aboveground herbivorous macroarthropods from ungrazed to lightly grazed treatments may be the result of greater quality of plant tissues in the lightly grazed treatment (Milchunas et al., 1995). Homoptera (primarily leafhoppers) are especially abundant in this treatment (Lavigne et al., 1972). Coleoptera (beetles, especially the most abundant family, Scarabaeidae) are generally relatively scarce in the heavily grazed treatment, and Chrysomelidae (leaf-feeding beetles) are found most commonly in the permanently ungrazed treatment. At the CPER,

d nd s ds un s ou d ro opod po gr ropo g o e r r th ov rth low rth ar Ab roa Be roa ro c c c i a a M M M

s

de

ato

m

Ne

Trophic composition (% of total)

100 80 60 40 20 0

U

H

U

H

L

H

U

M

Grazing intensity fungi

bacteria

herbivore

scavenger

omnivore

predator

Figure 18.3 Trophic structure of above- and belowground macroarthropods, microarthropods, and nematodes in relation to long-term grazing treatments in the northern shortgrass steppe. H, heavily grazed; L, lightly grazed; M, moderately grazed; U, ungrazed. (From Milchunas et al. [1998].)

470

Ecology of the Shortgrass Steppe

soil cores used to determine numbers of belowground macroarthropods included plant crowns and litter (Lloyd et al., 1973). Therefore, “soil” macroarthropods included some aboveground species, but did not include ants, which were dealt with separately. Two peaks, in May and then in August and September, in numbers and biomass of soil macroarthropods were observed in all treatments except the heavily grazed. In a separate 1-year study, Rottman and Capinera (1983) found smaller differences between grazing treatments in macroarthropod composition and/ or abundances than those reported earlier. Total biomass averaged 21, 35, and 17 mg · m–2 in the ungrazed, moderately, and heavily grazed treatments, respectively, with the moderately grazed value being significantly higher than the other two. Homoptera, Hemiptera, and Diptera followed patterns for abundance similar to that of biomass. However, total numbers of arthropods did not significantly differ among treatments. Western harvester ants and grasshoppers are conspicuous arthropods in the shortgrass steppe (Crist, chapter 10, this volume). Harvester ants can move 2.8 kg/colony of soil to the surface, and the disk around the mound cleared of vegetation can be more than 1 m in diameter (Rogers and Lavigne, 1974). Western harvester ant colonies are more numerous at intermediate intensities of long-term grazing, averaging 23, 28, 31, and 3 colonies/ha in the ungrazed, lightly, moderately, and heavily grazed treatments, respectively, at the CPER (Rogers et al., 1972). Disk diameters average 1.2, 0.9, 0.7, and 0.9 m in the previously noted grazing treatments, representing a range of only 0.02% to 0.3% of the land area cleared (Rogers and Lavigne, 1974). Differences between colonies in lightly and heavily grazed treatments are not apparent in terms of rates of forage extraction, foraging distance, time per foraging trip, or availability of seed (Rogers, 1974). Seed production of B. gracilis is lower in heavily grazed than ungrazed pastures (Coffin and Lauenroth, 1992), but Rogers (1974) observed that ants in the heavily grazed pasture tend to occupy areas where seeds are more abundant as a result of grazing of those areas late during the growing season. Seed consumption by ants is estimated to be only 2% of production. Grasshoppers normally constitute only 8% of total macroarthropod biomass (Lauenroth and Milchunas, 1991), but may exert significant pressure on plant communities during outbreak years, when populations are three to five times greater than average (Pfadt, 1977). Several studies have assessed grasshopper densities on the grazing treatments at the CPER. Van Horn et al. (1970) found no significant difference between lightly and heavily grazed treatments, but Capinera and Sechrist (1982) and Welch et al. (1991) found greater densities in lightly than in more heavily grazed treatments. Grasshoppers of different subfamilies responded differently to the grazing treatments. Catantopinae and Gomphocerinae were positively correlated with plant biomass (Capinera and Sechrist, 1982). Oedipodinae, generally associated with bare areas, were negatively correlated with plant biomass and thus increased with increasing grazing intensities, but were a smaller proportion of the total in all treatments. Studies conducted in mixed-grass and tallgrass prairies generally showed greater densities at higher grazing intensities (reviewed in Capinera and Sechrist, 1982).

Effects of Grazing on Abundance and Distribution of Consumers 471

Microarthropods In contrast to macroarthropods, microarthropods display only minor responses to long-term grazing treatments at the CPER (Leetham and Milchunas, 1985). Microarthropod sampling in this study did not include surface litter and crowns, and soil was sampled to a depth of 60 cm. Total numbers of microarthropods were 135,000 and 150,000 · m–2, and biomass totals were 62.3 and 68.7 mg · m–2, in the lightly and heavily grazed treatments, respectively. Differences between treatments in orders or families were not consistent with respect to trophic classification. However, significant increases with grazing were observed for Linotetranidae and Pseudococcidae, and decreases observed for Tardigrade and Bdellidae. The only significant effect of grazing intensity on depth distribution of microarthropod families was that Linotetranidae was found at deeper depths in the heavily grazed treatment. In contrast to grazing treatments, soil water and root biomass had large influences on microarthropods. Soil water was 19% greater in the heavily than lightly grazed treatment, which may have compensated for any negative effects of grazing. Root biomass did not differ with grazing treatment in this study. Crossley et al. (1975) sampled ungrazed, lightly, moderately, and heavily grazed treatments to a depth of only 5 cm and found no more than a “suggestion” of microarthropod abundance being greater in lightly and moderately grazed treatments. In this study also, grazing treatments had little effect on the composition of groups of microarthropods.

Nematodes and Microfungi Populations Plant parasitic nematodes are one of the three major groups of herbivores in the shortgrass steppe along with arthropods and large ungulates. This group of nematodes has been estimated to be 37% by Leetham and Scott (unpublished data), and 38% to 43% by Wall-Freckman and Huang (1998) of all nematodes in this system. Bacterial feeders, fungal feeders, predators, and omnivores are important regulators of decomposition and nutrient cycling in ecosystems (Freckman and Caswell, 1985). There is some evidence to suggest that herbivory by nematodes may exert a greater influence on primary production than would be predicted based upon their levels of consumption (Detling et al., 1980; Stanton, 1983). Leaf photosynthetic rates increased 35% when B. gracilis roots were clipped (Detling et al., 1980). Furthermore, aboveground herbivory can influence root chemistry (Cook et al., 1958; Kinsinger and Hopkins, 1961) and exudation (Dyer and Bokhari, 1976; Vancura and Stanek, 1975). Therefore, interactions may occur between aboveand belowground herbivores. Nematodes show very little response to grazing at the CPER. No significant differences were found in abundances of nematodes between long-term ungrazed and moderately grazed treatments, or between long-term grazing treatments or those recently exposed to grazing after long-term exclosure and recently exclosed to grazing after long-term moderate grazing (Wall-Freckman and Huang, 1998). Samples were collected both under plants (B. gracilis) and in the bare interspaces

472

Ecology of the Shortgrass Steppe

between them. Respective means were 1400 and 8500 nematodes · kg–1 dry soil in the ungrazed treatment and 1400 and 6800 nematodes · kg–1 dry soil in the grazed treatment. Bouteloua gracilis and total plant basal cover is greater in grazed treatments, which may tend to bring the means even closer. A total of 81 (under plant) and 83 (interplant) taxa were found in the ungrazed treatment compared with 82 and 79 in the grazed. No significant differences between long-term grazing treatments were found for species richness, fungivore/bacterivore, trophic diversity, Shannon-Weiner diversity (H⬘), exponential H⬘, Simpson’s diversity, evenness of the latter two, dominance diversity curves, two similarity indices, maturity index, or plant–parasite index. The proportion of bacterial feeders was slightly greater in long-term ungrazed than grazed treatment, and the opposite was found for plant parasites. Microfungal colonies averaged 84,500 and 93,000 · g–1 dry soil in the lightly and heavily grazed treatments, respectively (Christensen and Scarborough, 1969). An average of 64 species were identified in the lightly grazed treatment compared with 54 in the heavily grazed, and dominance of particular species increased with the decreasing species richness in the heavily grazed treatment. Species similarity was 73% between the two grazing treatments compared with 81% between lightly grazed replicates.

Diversity, Relative Sensitivity, and Trophic Structure of Consumer Groups The effect of grazing by domestic animals on biodiversity has been, and will probably continue to be, an important sociopolitical and conservation issue. Individual studies generally focus on one group of consumers and on its relationship to plant community or soil environment. Even when reviewed together, as in the previous sections, it is often difficult to discern relative sensitivities clearly. In this section we examine the relative degree to which long-term grazing by cattle has affected abundances, composition, and diversity by directly comparing differences in responses among the groups of organisms. Grazing generally has negative effects on the abundances of consumer groups of the shortgrass steppe (Fig. 18.4, [Milchunas et al., 1998]). The few cases in which increases occur are associated with lighter intensities of grazing. Groups of consumers respond very differently to grazing intensity. Aboveground macroarthropods have the greatest response, and rodents and microarthropods have the least. Dissimilarities in community composition of the groups of organisms (Fig. 18.5) do not always correspond to changes in abundance (Fig. 18.4) in relation to grazing intensities. Birds show small responses in overall abundance to grazing intensity, but are second to aboveground arthropods with respect to changes in community composition. Aboveground macroarthropods display large reductions in abundance with grazing, as well as large changes in species composition. Plants, microarthropods, and nematodes change little with grazing in either abundance or community composition.

ma abo cro veg art rou hr nd op od b m elo ac w ro gr ar ou th n ro d po m d icr oa rth ro ne po m d ato de

50

lag om or ph ro de nt bir d

pl an t

Effects of Grazing on Abundance and Distribution of Consumers 473

Abundance

(grazed % of ungrazed)

30 10 -10 -30 -50 -70 -90 U/L

U/M

U/H

L/M

L/H

Figure 18.4 The abundance (measured as grazed percent of ungrazed) of plants, lagomorphs, rodents, birds, above- and belowground macroarthropods, microarthropods, and nematodes in relation to long-term grazing treatment contrasts in the northern shortgrass steppe. H, heavily grazed; L, lightly grazed; M, moderately grazed; U, ungrazed. (From Milchunas et al. [1998].)

Bird and plant diversity decline with increasing intensity of grazing (Fig. 18.6). Most other groups either increase in diversity with increasing grazing intensity and/or display highest diversity at intermediate intensities of grazing. Macroarthropods increase and birds decrease in diversity with increasing intensity of grazing, whereas both groups have large changes in species composition. Birds are the only group that change dominant species from the lightly to the heavily grazed treatment (Milchunas et al., 1998). Over all indices, birds and aboveground macroarthropods appear to be the most sensitive to grazing in the northern shortgrass steppe. Changes in abundance and dissimilarity of communities are not related to responses in diversity, and there are no consistent responses in diversity between groups of organisms. Birds are the only group to display a similar response in diversity to that of plants. That diversity of an assemblage increases or decreases with grazing does not necessarily indicate whether the changes are qualitatively positive or negative. Plants and birds are perhaps the most studied groups of organisms, and the ones to which quality attributes such as endemic/exotic, globally rare/abundant, can be most readily applied to species. Our work has shown (Milchunas et al., chapter 16, this volume) that plant species decline in diversity may not be considered a negative

pl a

ma abov cro eg art rou hr nd op od b e m a lo cro wg art rou hr nd op mi od cro art hro ne po ma d tod e

(13 yrs pla ) nts (4 7y lag rs) om orp h rod en t bir d

Ecology of the Shortgrass Steppe

nt s

474

Species Dissimilarity

1.0

0.8

0.6

0.4

0.2

0

U/L

U/M

U/H

L/M

L/H

Figure 18.5 Species dissimilarity (Whittaker index of community association) of plants, lagomorphs, rodents, birds, above- and belowground macroarthropods, microarthropods, and nematodes in relation to long-term grazing treatment contrasts in the northern shortgrass steppe. H, heavily grazed; L, lightly grazed; M, moderately grazed; U, ungrazed. (From Milchunas et al. [1998].)

effect of grazing in this system that has evolved in the presence of large herbivores. Exotic and native weedy species contribute to the greater diversity of ungrazed compared with grazed communities. The situation for birds is somewhat less clear, possibly because migratory birds are subject to favorable or unfavorable conditions on both summer and winter ranges. Such factors as abundance, long-term population trends, and distributional characteristics, are used to determine vulnerability to extirpation (Carter and Barker, 1993) in bird species. In the shortgrass steppe, four out of six species of nesting birds have relatively high rankings of vulnerability (Table 18.1). Mountain Plovers, a federal category I species, are the most threatened grassland endemic species breeding in the shortgrass steppe. They nest primarily in heavily grazed areas. This is also true of McCown’s Longspur (Table 18.1). Horned Larks, which are widespread in North America and not threatened, also increase with grazing. Two other endemics found in the shortgrass steppe but more closely associated with mixed-grass prairie (Lark Bunting, Chestnut-collared Longspur) prefer lightly grazed shortgrass steppe, but are more abundant in moderately and heavily grazed areas in the more productive mixed-grass prairie (Kantrud and Kologiski, 1982). Knopf (1996) commented

s)

s)

13

yr

Diversity (exp H’)

2.0

P

47

t(

t(

lan

lan

yr

o

om

g La

P

h rp

t

en

d Ro

rd

Bi

d d d d po d po po unthro oun thro hro e o rt gr ar od gr ar ve cro low cro croa mat o a a i e Ab M Be M N M

1.7 1.4 1.1 0.8 0.5 0.2

Grazing Intensity Ungrazed

Light

Medium

Heavy

Figure 18.6 Species diversity (exp. H⬘) of plants, lagomorphs, rodents, birds, above- and belowground macroarthropods, microarthropods, and nematodes in relation to long-term grazing treatments in the northern shortgrass steppe. (From Milchunas et al. [1998].)

Table 18.1 Breeding Bird Population Response to Long-Term Grazing Intensity Treatments in the Northern Shortgrass Steppe, with Population Abundance, Trend, Distribution, and Vulnerability Parameters Breeding Pairs/20aca Grazing Intensity Speciesd Horned Lark, S Lark Bunting, E W. Meadowlark, S McCown’s Longspur, E Mountain Plover, E,R Chestnut-c. Longspur, E a

L

M

H

2.5 5.3 1.3 2.25 0 0.3

4.5 3.1 1.7 5 0 0.5

7.3 0 0 3.65 1.8 0

BBSb no./Route

BBSb Trend

N.Am.

N.Am.

24.7 27.3 42.2 4.7 0.4 9.9

–0.7f –2.1f –0.5 +7.3g –3.7g +0.4

Distributionc Breed, Winter N.Am.

11 43 21 54 54 44

Vulnerable Indexc Colorado

1.43 3.29 2.14 3.71 4 3.57

Giezentanner (1970). Compiled from Breeding Bird Survey (BBS) data (1966–1993) by Knopf (1996). c Carter and Barker (1993). d E, endemic grassland; R, rare, federal category I candidate; S, Secondary grassland. e Other studies indicate this specie prefers heavy grazed sites. f Significant at P = .05. g Significant at P = .001. Breeding and winter distribution rank: 1, very widespread; 2, widespread; 3, intermediate; 4, local; 5, very local. Vulnerability to extirpation: 1, very low; 2, low; 3, moderate; 4, high; 5, very high. H, heavy; L, light; M, moderate; N. Am., North America. (From Milchunas et al. [1998].) b

476

Ecology of the Shortgrass Steppe

that all endemic grassland species of the Great Plains evolved within the context of an intensively grazed landscape, but that current management of domestic animals creates a uniformly grazed situation rather than the mosaic of differentially grazed sites that may have been characteristic of the grazing patterns of native ungulate herds. Cattle, arthropods, and nematodes are the primary herbivores in the shortgrass steppe, each representing almost a third of the total herbivory. The presence or absence of cattle in this system has a large influence on the amount of plant material available to other herbivores, and potentially on the trophic structure of all consumers. Abundances of some groups of consumers are dramatically affected by grazing intensity treatments (Fig. 18.4). Although some statistically significant differences have been found for each particular group of consumers, different grazing intensities have very little overall effect on trophic structure of arthropods or nematodes (Fig. 18.3). This is also true for the herbivore component, which tends to increase rather than decrease with grazing by cattle. Herbivore groups consuming relatively small amounts of plant material in the shortgrass steppe also display minor shifts in trophic structure. Small-mammal biomasses in herbivore, carnivore, omnivore, and granivore categories vary little among grazing treatments in the shortgrass steppe compared with very large differences found in differentially grazed tallgrass prairie (Grant et al., 1982). The proportion of insect and seed consumed by birds in the southern shortgrass steppe varies little between grazed and ungrazed treatments, but differences are evident in more productive mixed-grass prairie (Wiens and Dyer, 1975). Differences in trophic structure between grazed and ungrazed shortgrass steppe appear to be very small across all groups of organisms, and this is similar to the relatively small changes in plant species composition.

Summary and Overview Grazing of public lands by domestic livestock is a controversial issue in the United States (Painter and Belsky, 1993) and will probably continue to be so. In a review of species endangerment patterns in the continental United States, Flather et al. (1994) listed grazing as a contributing factor to endangerment of 187 of a total of 667 threatened and endangered species. Grazing ranked high as a factor for plants (104 of 285 species), intermediate for birds (31 of 85) and insects (9 of 22), and low for reptiles (6 of 33), mammals (10 of 68), snails (4 of 13), and amphibians (4 of 11 species). Regions identified with high species endangerment are those with a short evolutionary history of grazing (southwestern United States and areas west of the Rocky Mountains). This analysis of grazing effects on other consumers in the shortgrass steppe does not imply large negative impacts of livestock grazing. In contrast, the small effects of grazing on consumer trophic structure hints at a degree of stability. Birds are a particularly sensitive group to grazing. However, the response of the bird community suggests a conclusion similar to that for plants: The removal of the domestic large herbivore may be considered a disturbance in this system.

Effects of Grazing on Abundance and Distribution of Consumers 477

This may be the result of the long history of grazing by bison, and the relatively high similarity between diets of bison and cattle. However, the restricted movement of domestic stock compared with the unfettered wandering over long distances and wide areas by wild ungulates may be an ecologically important difference between the two (Werger, 1977). Research that contrasts rotational with season-long cattle grazing systems has found no differences with respect to plant community response in either the shortgrass steppe or nearby northern mixed-grass prairie (Derner and Hart, 2007). The influence of different grazing systems on wild consumers has not been explored in the shortgrass steppe. However, the degree to which any particular grazing system mimics the past pattern of free-ranging bison will always be speculative, because nominal conditions are not well documented. In addition to the long history of grazing, the short stature of this semiarid grassland is an additional factor in the small effects of livestock grazing on both plant and consumer populations. Grazing by large herbivores directly affects canopy structure, but taller canopies are more affected by vegetation removal and trampling than shorter ones. Competition for light between grazed and ungrazed neighbors is more likely to be altered in a tall canopy than in a short one. Thus, grazing in productive systems with a long evolutionary history of grazing, such as the tallgrass prairie of North America or the Serengeti of Africa, often results in very large responses in consumer dynamics. In even greater contrast to the shortgrass steppe, grazing of domestic herbivores in productive systems with a short evolutionary history of grazing can be a massively destructive force, resulting in extinctions in and greatly altered functioning of systems such as those in Australia, New Zealand, and the southwestern United States. The shortgrass steppe appears to be among the most grazing-resistant ecosystems in the world.

References Allen, P. F., and B. Osborn. 1954. Tallgrass defeats prairie dogs. Soil Conservation 20:103–105. Andrews, R. M. 1977. A shortgrass prairie macroarthropod community: Organization, energetics and effects of grazing by cattle. U.S. International Biological Program, Grassland Biome preprint no. 208. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. Austin, D. D., and P. J. Urness. 1986. Effects of cattle grazing on mule deer diet and area selection. Journal of Range Management 39:18–21. Bock, C. E., and B. Webb. 1984. Birds as grazing indicator species in southeastern Arizona. Journal of Wildlife Management 48:1045–1049. Bohnam, C. D., and A. Lerwick. 1976. Vegetation changes induced by prairie dogs on shortgrass range. Journal of Range Management 29:221–225. Brown, J. H. 1973. Species diversity of seed-eating desert rodents in sand dune habitats. Ecology 54:775–787. Carter, M. F., and K. Barker. 1993. An interactive database for setting conservation priorities for western neotropical migrants, pp. 120–144. In: D. M. Finch and P. W. Stangel (eds.), Status and management of neotropical migratory birds. USDA Forest Service,

478

Ecology of the Shortgrass Steppe

general technical report RM-229. Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colo. Capinera, J. L., and T. S. Sechrist. 1982. Grasshopper (Acrididae)–host plant associations: Response of grasshopper populations to cattle grazing intensity. Canadian Entomologist 114:1055–1062. Christensen, M., and A. M. Scarborough. 1969. Soil microfungi investigations Pawnee site. U.S. International Biological Program, Grassland Biome technical report no. 23. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. Clary, W. P., and D. M. Beale. 1983. Pronghorn reactions to winter sheep grazing, plant communities, and topography in the Great Basin. Journal of Range Management 36:749–752. Clary, W. P., and R. C. Holmgren. 1982. Observations of pronghorn distribution in relation to sheep grazing on the Desert Experimental Range, pp. 581–592. In: J. M. Peek and P. D. Dalke (eds.), Wildlife–livestock relationships symposium. University of Idaho, Forest, Wildlife, and Range Experiment Station, Moscow, Idaho. Cody, M. L. 1985. Habitat selection in birds. Academic Press, Orlando, Fla. Coffin, D. P., and W. K. Lauenroth. 1992. Spatial variability in seed production of the perennial grass Bouteloua gracilis (H.B.K.) LAG. ex Griffiths. American Journal of Botany 79:347–353. Cook, C. W., L. A. Stoddart, and F. Kinsinger. 1958. Responses of crested wheat grass to various clipping treatments. Ecological Monographs 38:237–272. Coppock, D. L., J. K. Detling, J. E. Ellis, and M. I. Dyer. 1983. Plant–herbivore interactions in a North American mixed-grass prairie, I. Effects of black-tailed prairie dogs on intraseasonal aboveground plant biomass and nutrient dynamics and plant species diversity. Oecologia 56:1–9. Crossley, D. A., Jr., C. W. Proctor, Jr., and C. Gist. 1975. Summer biomass of soil microarthropods of the Pawnee National Grassland, Colorado. American Midland Naturalist 93:491–495. Cully, J. F., Jr. 1991. Response of raptors to reduction of a Gunnison’s prairie dog population by plague. American Midland Naturalist 125:140–149. Derner, J. D., J. K. Detling, and M. F. Antolin. 2006. Are livestock weight gains affected by black-tailed prairie dogs? Frontiers in Ecology and Environment 4:459–464. Derner, J. D., and R. H. Hart. 2007. Grazing-induced modifications to peak standing crop in northern mixed-grass prairie. Rangeland Ecology and Management 60:270–276. Detling, J. K., D. J. Winn, C. Procter-Gregg, and E. L. Painter. 1980. Effects of simulated grazing by belowground herbivores on growth, CO2 exchange, and carbon allocation patterns of Bouteloua gracilis. Journal of Applied Ecology 17:771–778. Dyer, M. I., and U. G. Bokhari. 1976. Plant–animal interactions: Studies of the effects of grasshopper grazing on blue grama grass. Ecology 57:762–772. Flake, L. D. 1971. An ecological study of rodents in a shortgrass prairie of northeastern Colorado. U.S. International Biological Program, Grassland Biome technical report no. 100. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. Flather, C. H., L. A. Joyce, and C. A. Bloomgarden. 1994. Species endangerment patterns in the United States. USDA Forest Service, general technical report RM-241. Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colo. Flinders, J. T., and R. M. Hansen. 1972. Diets and habits of jackrabbits in northeastern Colorado. Science series, 1. Range Science Department, Colorado State University, Fort Collins, Colo.

Effects of Grazing on Abundance and Distribution of Consumers 479 Flinders, J. T., and R. M. Hansen. 1975. Spring population responses cottontails and jackrabbits to cattle grazing shortgrass prairie. Journal of Range Management 28:290–293. Freckman, D. W., and E. P. Caswell. 1985. The ecology of nematodes in agroecosystems. Annual Review Phytopathology 23:275–296. French, N. R., W. E. Grant, W. Grodzinski, and D. M. Swift. 1976. Small mammal energetics in grassland ecosystems. Ecological Monographs 46:201–220. Giezentanner, J. B. 1970. Avian distribution and population fluctuations on the shortgrass prairie of north central Colorado. U.S. International Biological Program, Grassland Biome technical report no. 62. Natural Resource Ecology Laboratory, Fort Collins, Colo. Gordon, I. J. 1988. Facilitation of red deer grazing by cattle and its impact on red deer performance. Journal of Applied Ecology 25:1–10. Grant, W. E., and E. C. Birney. 1979. Small mammal community structure in North American grasslands. Journal of Mammalogy 60:23–36. Grant, W. E., E. C. Birney, N. R. French, and D. M. Swift. 1982. Structure and productivity of grassland small mammal communities related to grazing-induced changes in vegetation cover. Journal of Mammalogy 63:248–260. Grant, W. E., N. R. French, and L. J. Folse, Jr. 1980. Effects of pocket gopher mounds on plant production in shortgrass prairie ecosystems. Southwest Naturalist 25:215–224. Graul, W. D. 1973. Adaptive aspects of the Mountain Plover social system. Living Bird 12:69–94. Graul, W. D. 1975. Breeding biology of the Mountain Plover. Wilson Bulletin 87:6–31. Grzybowski, J. A. 1980. Ecological relationships among grassland birds during winter. PhD diss., University of Oklahoma, Norman, Okla. Grzybowski, J. A. 1982. Population structure in grassland bird communities during winter. Condor 84:137–152. Guenther, D. A., and J. K. Detling. 2003. Observations of cattle use of prairie dog towns. Journal of Range Management 56:410–417. Hailey, T. L., J. W. Thomas, and R. M. Robinson. 1966. Pronghorn die off in Trans-Pecos Texas. Journal of Wildlife Management 30:488–496. Hansen, R. M., and I. K. Gold. 1977. Blacktail prairie dogs, desert cottontails and cattle trophic relations on shortgrass range. Journal of Range Management 30:210–214. Hart, R. H. 1994. Rangeland, pp. 491–501. In: C. Arntzen (ed.). Encyclopedia of agricultural science, vol. 3. Academic Press, Los Angeles, Calif. Hassien, F. D. 1976. A search for black-footed ferrets in the Oklahoma Panhandle and adjacent area and an ecological study of black-tailed prairie dogs in Texas County, Oklahoma. Masters thesis, Oklahoma State University, Stillwater, Okla. Holland, E. A., and J. K. Detling. 1990. Plant responses to herbivory and belowground nitrogen cycling. Ecology 71:1040–1049. Hoover, R. L., C. E. Till, and S. Ogilvie. 1959. The antelope of Colorado. Colorado Game, Fish and Parks Department, technical bulletin no. 4. State of Colorado, Denver. Hornaday, W. T. 1889. The extermination of the American bison with a sketch of its discovery and life history. Smithsonian report 1887. Smithsonian Institution, Washington, D.C. Kantrud, H. A., and R. L. Kologiski. 1982. Effects of soils and grazing on breeding birds of uncultivated upland grasslands of the northern Great Plains. U.S. Department of Interior–Fish and Wildlife Service, Wildlife Research report 15. King, R. 1978. Habitat use and relative behaviors of breeding Long-billed Curlews. Masters thesis. Colorado State University, Fort Collins, Colo.

480

Ecology of the Shortgrass Steppe

Kinsinger, F. E., and H. H. Hopkins. 1961. Carbohydrate content of underground parts of grasses as affected by clipping. Journal of Range Management 14:9–12. Knopf, F. L. 1996. Prairie legacies: Birds, pp. 135–148. In: F. B. Samson and F. L. Knopf (eds.), Prairie conservation: Preserving North America’s most endangered ecosystem. Island Press, Covelo, Calif. Knowles, C. J. 1986. Some relationships of black-tailed prairie dogs to livestock grazing. Great Basin Naturalist 46:198–203. Koford, C. B. 1958. Prairie dogs, whitefaces and blue grama. Wildlife Monograph 3:1–78. Langner, L. L., and C. H. Flather. 1994. Biological diversity: Status and trends in the United States. USDA Forest Service, general technical report RM-244. Rocky Mountain Forest and Range Experiment Station, Fort Collins, Colo. Larson, F. 1940. The role of the bison in maintaining the short grass plains. Ecology 21:113–121. Lauenroth, W. K., and D. G. Milchunas. 1991. The shortgrass steppe, pp. 183–226. In: R. T. Coupland (ed.), Natural grasslands, introduction and western hemisphere. Ecosystems of the world 8A. Elsevier, Amsterdam. Lavigne, R. J., R. Kumar, J. Leetham, and V. Keith. 1972. Population densities and biomass of aboveground arthropods under various grazing and environmental stress treatments on the Pawnee site, 1971. U.S. International Biological Program, Grassland Biome technical report no. 204. Natural Resource Ecology Laboratory, Fort Collins, Colo. Leetham, J. W., and D. G. Milchunas. 1985. The composition and distribution of soil microarthropods in the shortgrass steppe in relation to the soil water, root biomass, and grazing by cattle. Pedobiologia 28:311–325. Lloyd, J. E., R. Kumar, R. R. Grow, J. W. Leetham, and V. Keith. 1973. Abundance and biomass of soil macroinvertebrates of the Pawnee site collected from pastures subjected to different grazing pressures, irrigation and/or water fertilization, 1970–1971. U.S. International Biological Program, Grassland Biome technical report no. 239. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. MacArthur, R. H. 1965. Patterns of species diversity. Biological Review 40:510–533. McNulty, F. 1971. Must they die? Doubleday, Garden City, N.J. Merriam, C. H. 1902. The prairie dog of the Great Plains, pp. 257–270. In: United States Department of Agriculture yearbook 1901. U.S. Government Printing Office, Washington, D.C. Milchunas, D. G., W. K. Lauenroth, and I. C. Burke. 1998. Livestock grazing: Animal and plant biodiversity of shortgrass steppe and the relationship to ecosystem function. Oikos 83:65–74. Milchunas, D. G., W. K. Lauenroth, P. L. Chapman, and M. K. Kazempour. 1989. Effects of grazing, topography, and precipitation on the structure of a semiarid grassland. Vegetatio 80:11–23. Milchunas, D. G., O. E. Sala, and W. K. Lauenroth. 1988. A generalized model of the effects of grazing by large herbivores on grassland community structure. American Naturalist 132:87–106. Milchunas, D. G., A. S. Varnamkhasti, W. K. Lauenroth, and H. Goetz. 1995. Forage quality in relation to long-term grazing history, current-year defoliation, and water resource. Oecologia 101:366–374. Miller, B., G. Ceballos, and R. Reading. 1994. The prairie dog and biotic diversity. Conservation Biology 8:677–681.

Effects of Grazing on Abundance and Distribution of Consumers 481 Morris, M. G. 1973. The effects of seasonal grazing on the Heteroptera and Auchenorrhyncha (Hemiptera) of chalk grassland. Journal of Applied Ecology 10:761–780. Olendorff, R. R. 1972. The large birds of prey of the Pawnee National Grassland: Nesting habits and productivity, 1969–1971. U.S. International Biological Program, Grassland Biome technical report no. 151. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. Olendorff, R. R., and J. W. Stoddart. 1974. The potential for management of raptor populations in western grasslands, pp. 44–88. In: F. N. Hammerstrom, B. E. Harrell, and R. R. Olendorff (eds.), Management of raptors. Raptor research report 2. Raptor Research Foundation, Vermillion, S.Dak. O’Meilia, M. E., F. L. Knopf, and J. C. Lewis. 1982. Some consequences of competition between prairie dogs and beef cattle. Journal of Range Management 35:580–585. Painter, E. L., and A. J. Belsky. 1993. Application of herbivore optimization theory to rangelands of the western United States. Ecological Applications 3:2–9. Paton, P. W. C., and J. Dalton. 1994. Breeding ecology of Long-billed Curlews at Great Salt Lake, Utah. Great Basin Naturalist 54:79–85. Peden, D. G., G. M. Van Dyne, R. W. Rice, and R. M. Hansen. 1974. The trophic ecology of Bison bison L. on shortgrass plains. Journal of Applied Ecology 11:489–497. Pfadt, R. E. 1977. Some aspects of the ecology of grasshopper populations inhabiting the shortgrass prairie, pp. 73–79. In: H. M. Kulman and H. C. Chiang (eds.), Insect ecology. Papers presented in the A. C. Hodson lectures. University Minnesota Agricultural Experiment Station technical bulletin 310. University of Minnesota, St. Paul. Phillips, P. 1936. The distribution of rodents in overgrazed and normal grasslands of central Oklahoma. Ecology 17:673–679. Plumb, G. E., and J. L. Dodd. 1993. Foraging ecology of bison and cattle on a mixed prairie: Implications for natural area management. Ecological Applications 3:631–643. Rebollo, S., D. G. Milchunas, I. Noy-Meir, and P. L. Chapman. 2002. The role of a spiny plant refuge in structuring grazed shortgrass steppe plant communities. Oikos 98:53–64. Rice, R. W., R. B. Dean, and J. E. Ellis. 1974. Bison, cattle, and sheep dietary quality and food intake. Proceedings of the Western Section of the American Society of Animal Science 25:194–197. Roberts, H. B., and K. G. Becker. 1982. Managing central Idaho rangelands for livestock and elk, pp. 537–543. In: J. M. Peek and P. D. Dalke (eds.), Wildlife–livestock relationships symposium. University of Idaho, Forest, Wildlife, and Range Experiment Station, Moscow, Idaho. Rogers, L. E. 1974. Foraging activity of the Western harvester ant in the shortgrass plains ecosystem. Environmental Entomology 3:420–424. Rogers, L. E., and R. J. Lavigne. 1974. Environmental effects of Western harvester ants on the shortgrass plains ecosystem. Environmental Entomology 3:994–997. Rogers, L. E., R. J. Lavigne, and J. L. Miller. 1972. Bioenergetics of the Western harvester ant in the shortgrass plains ecosystem. Environmental Entomology 1:763–768. Rosenzweig, M. L., B. Smigel, and A. Kraft. 1975. Patterns of food, space, and diversity, pp. 241–268. In: I. Prakash and P. K. Ghosh (eds.), Rodents in desert environments. Dr. W. Junk, The Hague, the Netherlands. Rottman, R. J., and J. L. Capinera. 1983. Effects of insect and cattle-induced perturbations on a shortgrass prairie arthropod community. Journal of the Kansas Entomological Society 56:241–252.

482

Ecology of the Shortgrass Steppe

Ryder, R. A. 1980. Effects of grazing on bird habitats, pp. 51–66. In: R. M. DeGraaf (ed.), Management of western forests and grasslands for nongame birds. USDA Forest Service general technical report INT-86. USDA Forest Service, Ogden, Utah. Sanderson, R. H. 1959. Relationship between jackrabbit use and availability of forage on native sandhills range. Masters thesis, Colorado State University, Fort Collins, Colo. Schmutz, E. M., E. L. Smith, P. R. Ogden, M. L. Cox, J. O. Klemmedson, J. J. Norris, and L. C. Fierro. 1992. Desert grassland, pp. 337–362. In: R. T. Coupland (ed.), Natural grasslands, introduction and western hemisphere. Ecosystems of the world 8A. Elsevier, Amsterdam. Schwartz, C. C., and J. E. Ellis. 1981. Feeding ecology and niche separation in some native and domestic ungulates on the shortgrass prairie. Journal of Applied Ecology 18:343–353. Schwartz, C. C., and J. G. Nagy. 1976. Pronghorn diets relative to forage availability in northeastern Colorado. Journal of Wildlife Management 40:469–478. Schwartz, C. C., J. G. Nagy, and R. W. Rice. 1977. Pronghorn dietary quality relative to forage availability and other ruminants in Colorado. Journal of Wildlife Management 41:161–168. Shackford, J. S. 1991. Breeding ecology of the Mountain Plover in Oklahoma. Bulletin of the Oklahoma Ornithological Society 24:9–13. Smith, C. C. 1940. The effect of overgrazing and erosion upon the biota of the mixedgrass prairie of Oklahoma. Ecology 21:381–397. Sparks, K. L. 1972. Grazing behavior of bison and cattle on a shortgrass prairie. U.S. International Biological Program, Grassland Biome technical report no. 149. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. Stanton, N. L. 1983. The effect of clipping and phytophagous nematodes on net primary production of blue grama, Bouteloua gracilis. Oikos 40:249–257. Taylor, W. P., and D. W. Lay. 1944. Ecologic niches occupied by rabbits in eastern Texas. Ecology 25:120–121. Taylor, W. P., C. T. Vorhies, and W. P. Lister. 1935. The relation of jack rabbits to grazing in southern Arizona. Journal of Forestry 33:490–498. Valone, T. J., and J. H. Brown. 1995. Effects of competition, colonization, and extinction on rodent species diversity. Science 267:880–883. Vancura, V., and M. Stanek. 1975. Root exudates of plants V. Kinetics of exudates from bean root as related to the presence of reserve compounds in cotyledons. Plant and Soil 43:547–559. Van Horn, D. H., R. Mason, P. Grove, D. Maujean, and G. Hemming. 1970. Grasshopper population numbers and biomass dynamics on the Pawnee site. U.S. International Biological Program, Grassland Biome technical report no. 148. Natural Resource Ecology Laboratory, Colorado State University, Fort Collins, Colo. Van Vuren, D. 1982. Comparative ecology of bison and cattle in the Henry Mountains, Utah, pp. 449–457. In: J. M. Peek and P. D. Dalke (eds.), Wildlife–livestock relationships symposium. University of Idaho, Forest, Wildlife, and Range Experiment Station, Moscow, Idaho. Wakeley, J. S. 1978. Factors affecting the use of hunting sites by Ferruginous Hawks. Condor 80:316–326. Wall-Freckman, D., and S. P. Huang. 1998. Response of the soil nematode community in a shortgrass steppe to long-term and short-term grazing. Applied Soil Ecology 9:39–44.

Effects of Grazing on Abundance and Distribution of Consumers 483 Welch, J. L., R. Redak, and B. C. Kondratieff. 1991. Effects of cattle grazing on the density and species of grasshoppers (Orthoptera: Acrididae) of the Central Plains Experimental Range, Colorado: A reassessment after two decades. Journal of the Kansas Entomological Society 64:337–343. Werger, M. J. A. 1977. Effects of game and domestic livestock on vegetation in east and southern Africa, pp. 149–159. In: W. Krause (ed.), Handbook of vegetation science, part XIII. Application of vegetation science to grassland husbandry. Dr. W. Junk Publishers, The Hague, the Netherlands. Whicker, A. D., and J. K. Detling. 1988. Ecological consequences of prairie dog disturbances. BioScience 38:778–785. Wiens, J. A. 1969. An approach to the study of ecological relationships among grassland birds. Ornithology Monographs 8:1–93. Wiens, J. A. 1973. Pattern and process in grassland bird communities. Ecological Monographs 43:237–270. Wiens, J. A. 1974a. Climatic instability and the “ecological saturation” of bird communities in North American grasslands. Condor 76:385–400. Wiens, J. A. 1974b. Habitat heterogeneity and avian community structure in North American grasslands. American Midland Naturalist 91:195–213. Wiens, J. A., and M. I. Dyer. 1975. Rangeland avifaunas: Their composition, energetics, and role in the ecosystem, pp. 146–182. In: D. R. Smith general technical report WO-1. USDA Forest Service, Washington, D.C. Wiens, J. A., and J. T. Rotenberry. 1979. Diet niche relationships among North American grassland and shrubsteppe birds. Oecologia 42:253–292. With, K. A. 1994. McCown’s Longspur. Birds of North America 96:8–22.

19 The Future of the Shortgrass Steppe Ingrid C. Burke William K. Lauenroth Michael F. Antolin Justin D. Derner Daniel G. Milchunas Jack A. Morgan Paul Stapp

W

here lies the future of the shortgrass steppe? In prior chapters we have described the remarkable resilience of the shortgrass steppe ecosystem and its organisms to past drought and grazing, and their sensitivity to other types of change. Emerging from this analysis is the idea of vulnerability to two main forces: future changes in precipitation or water availability, and direct human impacts. What are the likely changes in the shortgrass steppe during the next several decades? Which of the changes are most likely to affect major responses in the plants, animals, and ecosystem services of the shortgrass steppe? In this chapter we evaluate the current status of the shortgrass steppe and its potential responses to three sets of factors that will be driving forces for the future of the steppe: land-use change, atmospheric change, and changes in diseases.

Land-Use Change and Conservation Conservation and Management Challenges in the Shortgrass Steppe: Traditional and Emerging Land-Use Practices

Paul Stapp, Justin D. Derner, Ingrid C. Burke, William K. Lauenroth, Michael F. Antolin

484

The Future of the Shortgrass Steppe 485

Referring to the early 1900s, James Michener in his novel Centennial (1974) wrote the following: The old two-part system that had prevailed at the end of the nineteenth century— rancher and irrigator—was now a tripartite cooperation: the rancher used the rougher upland prairie; the irrigation farmer kept to the bottom lands; and the drylands gambler plowed the sweeping field in between, losing his seed money one year, reaping a fortune the next, depending on the rain. It was an imaginative system, requiring three different types of man, three different attitudes toward life. . . . (p. 1081)

Even today, because of the strong water limitation for cropping, the shortgrass steppe remains relatively intact, or at least unplowed, in contrast to other grassland ecosystems (Samson and Knopf, 1994). More than half of the shortgrass steppe remains in untilled, landscape-scale tracts, compared with only 9% of tallgrass prairie and 39% of mixed-grass prairie (The Nature Conservancy, 2003). These large tracts, including those in the national grasslands (Pawnee, Cimarron, Comanche, and Kiowa/Rita Blanca), provide the greatest opportunity for preserving key ecological processes and biological diversity. The landscape of the 1900s has been rapidly changing during the past several decades. Increased habitat loss and fragmentation threaten biological diversity in the shortgrass steppe. Land-use changes in the shortgrass steppe are similar to those throughout the United States, where, during the latter half of the 20th century, most human population growth in lower density regions surrounds urban centers, contributing to land-use shifts from agricultural to exurban developments (Brown et al., 2005; Theobald, 2005). During recent years, exurban land development increased at a rate 25 times higher than overall U.S. population growth (Theobald, 2005). During the next decade and a half, it is estimated that exurban developments will expand to 14.3% of U.S. land area. The result for eastern Colorado may be a loss from production of as much as 35% of row–crop agriculture (Parton et al., 2003). Concomitantly, the composition of the rural population is changing, as the proportion of elderly people on agricultural lands increases (Hauteniemi and Gutmann, 2005; Parton et al., 2007a). These changes are highly significant to agricultural ecosystem services (Millennium Ecosystem Assessment Series, 2003), including food, timber, and fiber production, as well as to cultural ecosystem services associated with open space, the social structure of rural communities, ecosystem processes including productivity, and biological diversity (Theobald, 2004). The interactions among land use, social processes and cultural values, and ecosystem services represent new frontiers in both ecological and social sciences (Gunderson et al., 2005), and pose many of the most important challenges for the future of the shortgrass steppe. In the eastern plains of Colorado, population has grown exponentially during the past 50 years (2.47% increase per year; data from U.S. Census Bureau [2000]), especially in the 11 highly developed counties bordering the Front Range of the Rocky Mountains (Fig. 19.1). Between 1987 and 2002, the number of housing units in these counties increased by 30%, to more than 1.5 million, compared with only a 2% increase in the more rural counties of the plains (data from

(A) Percent Change in Population (1950-2006)

(B) Percent Change in Housing Units (1950-2005)

Sedgewick Logan

Larimer

Sedgewick Logan

Larimer

Phillips

Weld

Morgan

Morgan

Boulder Denver

Yuma

Washington

Adams

Denver

Arapahoe

Jefferson Elbert

Elbert

Kit Carson

486

Teller

Cheyenne

El Paso

Fremont

Figure 19.1

Fremont

Crowley Otero

Las Animas

Cheyenne

El Paso Lincoln

Kiowa Pueblo

-53 to -25 -24 to 0 Huerfano 1 to 60 61 to 350 351 to 700 701 to 7417

Kit Carson

Douglas

Lincoln

% Change

Washington

Adams

Arapahoe

Douglas Teller

Boulder

Yuma

Jefferson

Phillips

Weld

Bent

Prowers

Baca

% Change -24 to -8 -7 to 0 Huerfano 1 to 20 21 to 300 301 to 700 701 to 6494

Kiowa Pueblo

Crowley Otero

Las Animas

Bent

Prowers

Baca

The changes in population (A), housing units (B), since 1950 for the available census data for eastern Colorado (U.S. Census).

(C) Percent Change in Number of Farms (1950-2002)

(D) Percent Change in Farms < 20 ha (1950-2002)

Sedgewick Logan

Larimer

Sedgewick Logan

Larimer

Phillips

Weld

Morgan

Morgan

Boulder Denver

Denver Jefferson

Elbert

Arapahoe Elbert

Kit Carson

Kit Carson

Douglas

487

Teller

Cheyenne

El Paso Lincoln

Fremont

-94 to -60 -59 to -37 -36 to -20 -19 to 0 1 to 50 51 to 120

Washington

Adams

Arapahoe

Douglas

Fremont

Crowley Otero

Bent

Prowers

Huerfano

Las Animas

Cheyenne

El Paso Lincoln

Kiowa Pueblo

% Change

Yuma

Washington

Adams

Teller

Boulder

Yuma

Jefferson

Phillips

Weld

Baca

% Change -87 to -77 -76 to -50 -49 to -22 -21 to 0 1 to 100 101 to 396

Kiowa Pueblo

Crowley Otero

Bent

Prowers

Huerfano

Las Animas

Baca

Figure 19.1 Numbers of farms (C), and numbers of farms less than 20 ha (D) since 1950 for the available census data for eastern Colorado (U.S. Census). The numbers of farms have increased because the average farm size is decreasing in the areas closest to urban growth along the Front Range of the Rockies.

488

Ecology of the Shortgrass Steppe

Colorado Department of Local Affairs [2004]). During the same period, eastern Colorado lost 0.7 million ha, or 7%, of its agricultural land to other uses, with significant losses (10%) occurring in rangeland and other pasturelands (data from U.S. Census of Agriculture [2004]). Much of the development along the Front Range has occurred as rural residential or exurban development that fragments larger agricultural lands into smaller parcels, particularly the richest bottomlands formerly devoted to crops. In counties along the Interstate Highway 25 corridor, both the total number of farms and the number of small farms (